Intelligent water pipe processing method and system based on BIM (Building Information Modeling)

Abstract

The invention relates to the technical field of BIM (Building Information Modeling) water pipe processing, in particular to an intelligent water pipe processing method and system based on BIM. Identifying abnormal characteristics of abnormal historical water pipe data through a water pipe abnormal matching model and mapping the abnormal characteristics to second pipeline data; calling an abnormal laying processing scheme in the water pipe laying processing schemes to obtain a first abnormal parameter, performing scheme optimization on the laying processing scheme of the third pipeline data through the particle swarm optimization model and the first abnormal parameter, and performing water pipe processing according to the optimized scheme; through the method, data mapping can be carried out on the existing water pipe line data according to the historical water pipe line data with problems, the water pipe line data with possible problems are obtained to optimize the water pipe machining scheme, waste of materials and cost in the water pipe machining process is avoided, and the problem that current machining management is difficult is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of BIM water pipe processing, and in particular to a BIM-based water pipe intelligent processing method and system. Background Art

[0002] BIM technology is a digital management method in the construction field. By converting buildings into three-dimensional virtual models, it can quantify and visualize data for each construction step before construction, so as to formulate corresponding construction plans and management plans and improve the efficiency of subsequent construction.

[0003] Water pipes are an indispensable part of a building. The layout of the pipes not only affects the normal use of water facilities in the building after completion, but also has a certain impact on the subsequent progress of the construction process.

[0004] In the actual water pipe construction process, it is easy to have problems such as the water pipe is too long, the water pipe laying is blocked by other pipelines, the water pipe laying is blocked by other construction materials of the building, etc., which requires the water pipe to be processed on site. Due to the poor accuracy of the on-site secondary processing, the water pipe may be cut too short or burst, which can easily cause a large amount of waste of water pipe materials and increase costs, making the processing and management of water pipes more difficult.

[0005] Therefore, a water pipe intelligent processing method and system based on BIM is proposed. Summary of the invention

[0006] The purpose of the present invention is to provide a water pipe intelligent processing method and system based on BIM, which obtains the second pipeline data and the water pipe laying processing plan of the water pipe line data in the BIM model data, marks the historical water pipe line data according to the abnormal line points of the historical water pipe line data to generate the first pipeline data; identifies the abnormal characteristics of the first pipeline data through the water pipe abnormal matching model and marks the second pipeline data; and calls the abnormal laying processing plan corresponding to the first pipeline data in the water pipe laying processing plan to obtain the first abnormal parameter, optimizes the laying processing plan of the third pipeline data through the particle swarm optimization model and the first abnormal parameter, and processes the water pipe according to the optimized plan; through this method, the existing water pipe line data can be mapped according to the historical water pipe line data with problems, and the water pipe line data that may have problems can be obtained to optimize the water pipe processing plan, so as to avoid the waste of materials and costs in the water pipe processing process, so as to alleviate the current processing management difficulties.

[0007] To achieve the above object, the present invention provides the following technical solutions: A water pipe intelligent processing method based on BIM, comprising: Calling the BIM model data of the building, and obtaining the water pipe line data and the water pipe laying processing plan according to the BIM model data; Acquire abnormal line points of the water pipe line data of the historical BIM model data, mark the historical water pipe line data according to the abnormal line points to generate first pipeline data; generate second pipeline data according to the water pipe line data; Further, a three-dimensional coordinate system is established according to the BIM model data of the input water pipeline line, wherein the input water pipeline line includes the water pipeline line data and the historical water pipeline line data; the historical water pipeline line is mapped to the three-dimensional coordinate system to generate input water pipeline line coordinates; the input water pipeline line coordinates are segmented to generate an input water pipeline line coordinate set; According to the mapping of the three-dimensional coordinate system into a two-dimensional coordinate system, the coordinates of the input water pipe line coordinate set are projected into the two-dimensional coordinate system to generate an input water pipe line transposed coordinate set; according to the coordinates of the input water pipe line transposed coordinate set, a plurality of input water pipe line curves are generated; the input water pipe line curve includes the second pipeline data and the historical water pipe line; If the input water pipeline route is the historical water pipeline route, the coordinates of the abnormal route point marked by the historical water pipeline route coordinates are transposed into abnormal route transposition points, and multiple segments of the input water pipeline route curves are marked according to the abnormal route transposition points to generate the first pipeline data; Furthermore, the abnormal line transposition point includes multiple groups of data, and every two abnormal line transposition points form a group of data; Establishing a water pipe anomaly matching model, matching the data anomaly features of the first pipeline data according to the second pipeline data, marking the second pipeline data according to the anomaly matching result, and generating third pipeline data and fourth pipeline data; The water pipe anomaly matching model includes a line data compression unit, a line feature extraction unit, a line feature classification unit and an abnormal line output unit; The line data compression unit performs data standardization and data encoding on the second pipeline data; The line feature extraction unit extracts data features of the standardized second pipeline data; the line feature extraction unit includes a residual convolution network, an LSTM network and a feature fusion layer, the residual convolution network performs data feature enhancement and morphological feature analysis on the second pipeline data, the LSTM network extracts trend features of the second pipeline data, and the feature fusion layer fuses data features of the residual convolution network and the LSTM network; Furthermore, the number of the residual convolutional networks is 2, the number of the LSTM networks is 6, the feature fusion layer is Relu, and the line feature classification unit includes a fully connected layer and a Bayesian classifier; The line feature classification unit performs feature classification on the second pipeline data according to the feature label of the first pipeline data, and matches the feature label of the first pipeline data according to the classification result; The abnormal line output unit generates and outputs a feature label of the second pipeline data; Acquire a first target parameter according to the water pipe laying processing plan; extract the laying processing plan of the third pipeline data in the water pipe laying processing plan; call the abnormal laying processing plan of the corresponding first pipeline data according to the third pipeline data, and extract a first abnormal parameter according to the abnormal laying processing plan; Establishing a particle swarm optimization model, optimizing the laying and processing plan of the third pipeline data according to the first target parameter and the first abnormal parameter, and generating an abnormally optimized water pipe laying and processing plan; Further, a priority weight is set for the first target parameter according to the water pipe laying processing scheme, and the speed and position of the particle swarm are initialized with the first target parameter; the priority weight and the target first abnormal parameter are used as penalty coefficients, and the parameters of the water pipe laying processing scheme are used as target conditions, and the position and speed of the particle swarm are iterated multiple times until the particle swarm reaches the parameters of the water pipe laying processing scheme, and the abnormal optimized water pipe laying processing scheme is generated according to the first target parameter corresponding to the optimized particle swarm; Further, the priority weight is the cost weight and the length weight, the first target parameter is the number of water pipes and the length of water pipes; the target first abnormal parameter includes the number of abnormal water pipes and the length of abnormal water pipes; the parameter of the water pipe laying processing plan is the total length of the third pipeline data to be optimized; Processing the water pipes of the third pipeline data according to the abnormally optimized water pipe laying processing plan, and processing the fourth pipeline data according to the water pipe laying processing plan; The present invention also proposes a BIM-based water pipe intelligent processing system, including a water pipe line acquisition module, an abnormal line acquisition module, an abnormal line matching module, an abnormal parameter extraction module, an abnormal line optimization module and a normal and abnormal line processing plan generation module; specifically: A water pipe line acquisition module is used to acquire BIM model data of a building, and acquire water pipe line data and a water pipe laying and processing plan based on the BIM model data; and generate second pipeline data based on the water pipe line data; An abnormal line acquisition module is used to acquire abnormal line points of the water pipe line data of the historical BIM model data, and to generate first pipeline data according to marking the historical water pipe line data with the abnormal line points; An abnormal line matching module is used to establish a water pipe abnormal matching model, match the data abnormality characteristics of the first pipeline data according to the second pipeline data, mark the second pipeline data according to the abnormal matching result, and generate third pipeline data and fourth pipeline data; Furthermore, the water pipe anomaly matching model includes a line data compression unit, a line feature extraction unit, a line feature classification unit and an abnormal line output unit, wherein the line feature extraction unit includes 2 residual convolutional networks, 6 LSTM networks and 1 Relu feature fusion layer; the line feature classification unit includes a fully connected layer and a Bayesian classifier; The abnormal parameter extraction module obtains the first target parameter according to the water pipe laying processing plan; extracts the laying processing plan of the third pipeline data in the water pipe laying processing plan; calls the abnormal laying processing plan of the corresponding first pipeline data according to the third pipeline data, and extracts the first abnormal parameter according to the abnormal laying processing plan; The abnormal line optimization module establishes a particle swarm optimization model, optimizes the laying and processing plan of the third pipeline data according to the first target parameter and the first abnormal parameter, and generates an abnormal optimized water pipe laying and processing plan; A normal-abnormal line processing scheme generating module processes the water pipes of the third pipeline data according to the abnormal optimized water pipe laying processing scheme, and processes the fourth pipeline data according to the water pipe laying processing scheme; Furthermore, the abnormal line optimization module includes: setting a priority weight for the first target parameter based on the water pipe laying processing plan, and initializing the speed and position of the particle swarm with the first target parameter; using the priority weight and the target first abnormal parameter as a penalty coefficient, and the parameters of the water pipe laying processing plan as target conditions, iterating the position and the speed of the particle swarm multiple times until the particle swarm reaches the parameters of the water pipe laying processing plan, and generating the abnormal optimized water pipe laying processing plan based on the first target parameter corresponding to the optimized particle swarm.

[0008] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention extracts historical water pipe line data and water pipe line data through BIM model data, reduces the dimension of the data through the established three-dimensional coordinates, generates a two-dimensional curve and marks the abnormal line points of the historical water pipe line data. This method can reduce the original data dimension, and can perform data feature matching of the water pipe line through the two-dimensional data, thereby reducing the amount of data calculation from the data dimension. At the same time, it can also provide more accurate data for the subsequent third pipeline data matching, improve the accuracy of the third pipeline data matching, and provide a data basis for the subsequent water pipe processing plan.

[0009] 2. The present invention adopts a water pipe anomaly matching model and a method of time series data anomaly analysis. By converting the actual line laying problem into a time series anomaly classification problem, the classification matching degree of the water pipe line data after dimensionality reduction and the historical third pipeline data is analyzed. Through this method, whether there are lines in the current building water pipe lines that are difficult to lay with the historical water pipe lines can be matched, so as to determine the locations of water pipe lines that may be difficult to lay in the current plan, facilitate subsequent plan optimization and analysis, and avoid the same problem to provide more accurate processing plans to solve the problem of difficult processing management.

[0010] 3. The present invention is based on a particle swarm algorithm, based on mining outliers on water pipe lines that may be difficult to lay, by using water pipe processing parameters as particle groups, the weight of cost data and the parameters of water pipes that have had the same difficulty in laying as penalty functions, and the cost and number of water pipes in the laying and processing plan as optimization targets for optimization. This method can be based on the various parameters of the difficulty in laying water pipes that have been mined as empirical data, and at the same time combined with cost data, re-plan the difficult part of the laying plan for the existing plan to avoid the same water pipe laying difficulty problem, while reducing costs to form new plans to solve the problem of processing management difficulties. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 It is a schematic diagram of the method flow of the present invention; Figure 2 A schematic diagram of mapping the three-dimensional coordinate axis of the present invention into a two-dimensional coordinate axis; Figure 3 This is a schematic diagram of BIM water pipe line mapping time series data of the present invention; Figure 4 It is a flow chart of the water pipe abnormal matching model of the present invention; Figure 5 It is a schematic diagram of the accuracy of the training set and the test set of the water pipe anomaly matching model of the present invention; Figure 6 A schematic diagram of the loss function of the training set and the test set of the water pipe anomaly matching model of the present invention; Figure 7 It is a schematic diagram of the system flow of the present invention. DETAILED DESCRIPTION

[0012] 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.

[0013] In the process of laying water pipes according to the BIM model, various problems may occur during the construction process, resulting in the water pipes being unable to be laid according to the predetermined laying and processing plan. Workers will cut these water pipes for secondary processing, and in the secondary processing, it is easy to increase the materials and costs of water pipes due to problems such as worker mistakes. The length and thickness of the water pipes are managed according to the plan, but because the above problems often occur, it also leads to difficulties in the management of water pipe processing according to the plan. Therefore, a water pipe intelligent processing method based on BIM is proposed. Figure 1 As shown, the technical solution is as follows: Calling the BIM model data of the building, and obtaining the water pipe line data and the water pipe laying processing plan according to the BIM model data; Acquire abnormal line points of the water pipe line data of the historical BIM model data, mark the historical water pipe line data according to the abnormal line points to generate first pipeline data; generate second pipeline data according to the water pipe line data; Establishing a water pipe anomaly matching model, matching the data anomaly features of the first pipeline data according to the second pipeline data, marking the second pipeline data according to the anomaly matching result, and generating third pipeline data and fourth pipeline data; Acquire a first target parameter according to the water pipe laying processing plan; extract the laying processing plan of the third pipeline data in the water pipe laying processing plan; call the abnormal laying processing plan of the corresponding first pipeline data according to the third pipeline data, and extract a first abnormal parameter according to the abnormal laying processing plan; Establishing a particle swarm optimization model, optimizing the laying and processing plan of the third pipeline data according to the first target parameter and the first abnormal parameter, and generating an abnormally optimized water pipe laying and processing plan; The water pipes of the third pipeline data are processed according to the abnormally optimized water pipe laying processing plan, and the fourth pipeline data are processed according to the water pipe laying processing plan.

[0014] The water pipe line data with problems in the historical water pipe laying and processing plan are marked as abnormal, and the abnormal data with problems are mapped to the water pipe line data of the current plan through data mapping. In order to achieve this goal, the characteristics of each water pipe line data after abnormal marking are learned through the water pipe anomaly matching model, and the water pipe line data characteristics of the current plan are extracted. The water pipe lines with the same characteristics in the current plan are identified by comparing and classifying the two data features; this method can automatically identify and classify the water pipe lines with the same design as the water pipe lines with problems in the current plan through the model, so as to facilitate the subsequent optimization of the processing plan and reduce the difficulty of water pipe processing management.

[0015] Based on identifying the water pipe lines of the current solution with the same characteristics as the problem solution, the corresponding parameters of the water pipe laying and processing solution with the problem are called through the solution, and the corresponding parameters of the water pipe lines of the current solution with the same characteristics are used. Through the particle swarm model, the parameters of the water pipe laying and processing solution and the problematic water pipe lines are used as penalty coefficients and target conditions to optimize the parameters of the water pipe lines of the current solution. This method can optimize the data of the existing solution parameters, avoid the same solution parameters as historical problems according to the optimized solution, avoid the recurrence of the same problem, and reduce the difficulty of water pipe processing management.

[0016] Embodiment 1 For specific explanation, the following embodiments are used for elaboration: Acquire BIM model data of the building, and acquire water pipe line data and a water pipe laying processing plan according to the BIM model data; acquire abnormal line points of the water pipe line data of the historical BIM model data, and generate first pipeline data according to marking the historical water pipe line data with abnormal line points; generate second pipeline data according to the water pipe line data; the first pipeline data is the historical abnormal water pipe line data, and the second pipeline data is the target water pipe line data; the first pipeline data and the second pipeline data are both water pipe routing data of the water pipe in the BIM model; Furthermore, a three-dimensional coordinate system is established based on the BIM model data of the input water pipe line, and the input water pipe line includes the water pipe line data and the historical water pipe line data; the historical water pipe line is mapped to the three-dimensional coordinate system to generate the input water pipe line coordinates; the input water pipe line coordinates are segmented to generate the input water pipe line coordinate set; the segmentation rules of the input water pipe line coordinate segmentation include the specific usage content of the water pipe and the water pipe length determination, for example, when analyzing the sewer pipe, all the sewer line data are extracted; for example, when performing data analysis, the line data will be segmented to facilitate subsequent feature extraction, and the data segmentation will be divided according to the processing length of the water pipe during processing. For example, if it is processed according to 6 meters, it will be segmented according to integer multiples of 6, and if it is processed according to 4 meters and 6 meters, it will be divided according to the least common multiple of 6 and 4; according to the mapping of the three-dimensional coordinate system to a two-dimensional coordinate system, the coordinates of the input water pipe line coordinate set are projected into the two-dimensional coordinate system to generate the input water pipe line transposed coordinate set; the three-dimensional coordinate system is mapped to the two-dimensional coordinate system reference Figure 2 The specific calculation is as follows: ; in, The three-dimensional coordinate system is mapped to the x-axis of the two-dimensional coordinate system, x is the x-axis of the three-dimensional coordinate system, and z is the z-axis of the three-dimensional coordinate system. The three-dimensional coordinate system is mapped to the y-axis of the two-dimensional coordinate system, and y is the y-axis of the three-dimensional coordinate system. For x and The mapping angle of Figure 2 In As shown, is y and The mapping angle of Figure 2 In As shown; and The size of is set according to the specific implementation situation. The direction is consistent with the direction of water flow in the water pipe during actual use after the water pipe is laid; the purpose of the consistent water flow direction is to map the water flow direction into the time direction in the time series data; the purpose of this method is to determine the positions of the x-axis and the z-axis in the three-dimensional coordinates by the water flow direction, so as to unify the data and facilitate unified data mapping, avoiding data pollution caused by the different positions of the origin of the coordinate system and each coordinate axis; refer to Figure 3 As shown, Figure 3 A is the three-dimensional water pipe line of the BIM model. Figure 3 B is the one-dimensional water pipe line in the two-dimensional coordinate after mapping, where Figure 3 The solid thick line in B is the water pipe line data after mapping; Generate multiple segments of input water pipeline line curves according to the coordinates of the input water pipeline line transposed coordinate set; the input water pipeline line curve includes the second pipeline data and the historical water pipeline line; If the input water pipeline route is the historical water pipeline route, the coordinates of the abnormal route point marked by the historical water pipeline route coordinates are transposed into abnormal route transposition points, and multiple segments of the input water pipeline route curves are marked according to the abnormal route transposition points to generate the first pipeline data; By converting the three-dimensional coordinate data of the BIM model into two-dimensional coordinate data, and based on the characteristics of the water pipe line and its two-dimensional coordinate data characteristics, it becomes one-dimensional linear data; this method can reduce the feature dimension of the water pipe data of the BIM model, and at the same time, combine the characteristics of the data itself to convert the data into one-dimensional data. This method can reduce the feature dimension of the original data and ensure the maximum data characteristics of the data; at the same time, the data can also facilitate the subsequent feature analysis of the water pipe data, so as to provide an accurate and reliable data basis for the subsequent solution optimization; Further, the abnormal line transposition point includes multiple groups of data, and every two abnormal line transposition points form a group of data; the abnormal line transposition points are marked according to expert experience, and the two abnormal line transposition points of each group of data are the coordinate starting point of the historical abnormal water pipe line and the coordinate end point of the historical abnormal water pipe line; wherein the setting of the coordinate starting point and the coordinate end point is also divided according to the processing length of the water pipe during processing, for example, the setting is 6 meters for processing with 6 meters; the water pipe line data will be marked as 0 for normal data and 1 for abnormal data for feature marking; specifically, different types of data labels can be marked according to the specific implementation situation; By taking every two abnormal line transposition points as a group of data, the abnormal water pipes are marked, which facilitates the extraction of abnormal water pipe line data, and also facilitates unified data management and data label marking of these abnormal data, which facilitates the classification and identification of subsequent models and data feature matching, and provides an accurate and reliable data basis for subsequent solution optimization; Establishing a water pipe anomaly matching model, matching the data anomaly features of the first pipeline data according to the second pipeline data, marking the second pipeline data according to the anomaly matching result, and generating third pipeline data and fourth pipeline data; the third pipeline data is the abnormal water pipe line, and the fourth pipeline data is the normal water pipe line; The water pipe anomaly matching model includes a line data compression unit, a line feature extraction unit, a line feature classification unit and an abnormal line output unit. Figure 4 As shown; The line data compression unit performs data standardization and data encoding on the second pipeline data; the data standardization adopts Z-score standardization to compress the data to between 0 and 1, so as to retain the data form while reducing the interference of the data coordinate size on the recognition; the data encoding adopts one-hot encoding; The line feature extraction unit extracts data features of the standardized second pipeline data; the line feature extraction unit includes a residual convolution network, an LSTM network and a feature fusion layer, the residual convolution network performs data feature enhancement and morphological feature analysis on the second pipeline data, the LSTM network extracts trend features of the second pipeline data, and the feature fusion layer fuses data features of the residual convolution network and the LSTM network; Furthermore, the number of the residual convolutional networks is 2, the number of the LSTM networks is 6, the feature fusion layer is Relu, and the line feature classification unit includes a fully connected layer and a Bayesian classifier; the morphological abnormality features of the time series data converted from the water pipe line are extracted through the residual convolutional network, and the accuracy of abnormal feature recognition is improved by being able to mark key abnormal features; the trend features of the water pipe line are identified by combining the advantages of LSTM in the trend features of time series data, and anomaly detection is performed in combination with the overall change trend of the water pipe line, and the similar features of the current water pipe line and the historical water pipe line are fully identified from the spatial morphological dimension and the trend feature dimension, so as to improve the accuracy of abnormal water pipe line recognition; In this example, 80,000 water pipe line data were annotated by experts, and the training set and test set were divided into 8:2 ratio for training. The number of iterations was 100. The actual training results of the water pipe anomaly matching model are shown in Figure 2. Figure 5 and Figure 6 As shown, Figure 5 is the accuracy data of the training set and the test set, Figure 6 is the loss function of the training set and the test set, and the loss function adopts the L1 loss function; in this embodiment, some pipeline data that did not participate in the training are also annotated in combination with expert experience. There are a total of 5 groups of data, each group of data includes 100 water pipe lines, and the water pipe data includes abnormal lines and normal lines. The model of this embodiment is used to simulate the real environment for identification. The specific identification results are shown in Table 1: Table 1 Accuracy of water pipe anomaly matching model on abnormal lines and normal lines

[0017] From Table 1, we can see that although the accuracy of abnormality recognition is only 85% to 86%, the accuracy of abnormality recognition is as high as over 90%, and the accuracy of abnormality recognition is relatively high; The line feature classification unit performs feature classification on the second pipeline data according to the feature label of the first pipeline data, and matches the feature label of the first pipeline data according to the classification result; the abnormal line output unit generates and outputs the feature label of the second pipeline data; By establishing a water pipe anomaly matching model, the water pipe line data after dimensionality reduction is classified as time series data. According to the recognition method of the time series data anomaly recognition model, the abnormal water pipe line data is classified as abnormal time series data, and the current water pipe line data is classified as unrecognized time series data. The feature similarity between the current water pipe line data and the abnormal water pipe line data is detected in combination with the time series data anomaly detection method, so that the water pipe line in the current water pipe line data that is the same as the historical abnormal water pipe line data can be identified; through this method, the water pipe line can be automatically detected by the time series data detection method, so as to realize the automatic identification of abnormal water pipe lines, provide a water pipe line foundation for the optimization of subsequent solutions, facilitate the call of corresponding parameters to realize solution optimization, and reduce the difficulty of water pipe processing management; Acquire the first target parameter according to the water pipe laying processing plan; extract the laying processing plan of the third pipeline data in the water pipe laying processing plan; call the abnormal laying processing plan of the corresponding first pipeline data according to the third pipeline data, and extract the first abnormal parameter according to the abnormal laying processing plan; the first abnormal parameter is the water pipe parameter of the historical abnormal water pipe line data; the first target parameter is the water pipe parameter of the target water pipe line data; Establishing a particle swarm optimization model, optimizing the laying and processing plan of the third pipeline data according to the first target parameter and the first abnormal parameter, and generating an abnormally optimized water pipe laying and processing plan; Further, a priority weight is set for the first target parameter according to the water pipe laying processing scheme, and the speed and position of the particle swarm are initialized with the first target parameter; the priority weight and the target first abnormal parameter are used as penalty coefficients, and the parameters of the water pipe laying processing scheme are used as target conditions, and the position and speed of the particle swarm are iterated multiple times until the particle swarm reaches the parameters of the water pipe laying processing scheme, and the abnormal optimized water pipe laying processing scheme is generated according to the first target parameter corresponding to the optimized particle swarm; Through the particle swarm optimization algorithm, the corresponding parameters of the abnormal water pipe lines extracted from the historical data are used as penalty coefficients to avoid the same parameter data in the current plan. The corresponding parameters of the water pipe lines that are the same as the abnormal water pipe lines in the historical data are extracted as the initial data of the particle swarm algorithm, and the parameters of the water pipe laying and processing plan are used as the target conditions for optimization. This method can avoid the occurrence of the same problems under the constraints of the existing plan to reduce the difficulty of water pipe processing management; Furthermore, the priority weights are cost weights and length weights, the first target parameters are the number of water pipes and the length of water pipes; the target first abnormal parameters include the number of abnormal water pipes and the length of abnormal water pipes; the parameters of the water pipe laying and processing scheme are the total length of the third pipeline data to be optimized; the cost data of the water pipes is used as a penalty coefficient to form a constraint condition, and the scheme content is used as the optimization target, so that the cost investment of the water pipes can be reduced during the optimization process, and the optimization result will not exceed the content of the original water pipe design scheme, so as to improve the quality and reliability of the optimization result, and further alleviate the problem of difficult water pipe processing management; The water pipes of the third pipeline data are processed according to the abnormally optimized water pipe laying processing plan, and the fourth pipeline data are processed according to the water pipe laying processing plan.

[0018] Embodiment 2 The present invention also proposes a BIM-based water pipe intelligent processing system, including a water pipe line acquisition module, an abnormal line acquisition module, an abnormal line matching module, an abnormal parameter extraction module, an abnormal line optimization module and a normal abnormal line processing plan generation module. Figure 7 As shown; specifically: A water pipe line acquisition module is used to acquire BIM model data of a building, and acquire water pipe line data and a water pipe laying and processing plan based on the BIM model data; and generate second pipeline data based on the water pipe line data; An abnormal line acquisition module is used to acquire abnormal line points of the water pipe line data of the historical BIM model data, and to generate first pipeline data according to marking the historical water pipe line data with the abnormal line points; An abnormal line matching module is used to establish a water pipe abnormal matching model, match the data abnormality characteristics of the first pipeline data according to the second pipeline data, mark the second pipeline data according to the abnormal matching result, and generate third pipeline data and fourth pipeline data; Furthermore, the water pipe anomaly matching model includes a line data compression unit, a line feature extraction unit, a line feature classification unit and an abnormal line output unit, wherein the line feature extraction unit includes 2 residual convolutional networks, 6 LSTM networks and 1 Relu feature fusion layer; the line feature classification unit includes a fully connected layer and a Bayesian classifier; The abnormal parameter extraction module obtains the first target parameter according to the water pipe laying processing plan; extracts the laying processing plan of the third pipeline data in the water pipe laying processing plan; calls the abnormal laying processing plan of the corresponding first pipeline data according to the third pipeline data, and extracts the first abnormal parameter according to the abnormal laying processing plan; The abnormal line optimization module establishes a particle swarm optimization model, optimizes the laying and processing plan of the third pipeline data according to the first target parameter and the first abnormal parameter, and generates an abnormal optimized water pipe laying and processing plan; A normal-abnormal line processing scheme generating module processes the water pipes of the third pipeline data according to the abnormal optimized water pipe laying processing scheme, and processes the fourth pipeline data according to the water pipe laying processing scheme; Further, the abnormal line optimization module includes: setting a priority weight for the first target parameter according to the water pipe laying processing scheme, and initializing the speed and position of the particle swarm with the first target parameter; using the priority weight and the target first abnormal parameter as a penalty coefficient, and using the parameters of the water pipe laying processing scheme as target conditions, iterating the position and the speed of the particle swarm for multiple times until the particle swarm reaches the parameters of the water pipe laying processing scheme, and generating the abnormal optimized water pipe laying processing scheme according to the first target parameter corresponding to the optimized particle swarm; In this embodiment, some historical data that have not been trained are used for processing optimization simulation, and the final solution in the historical data and the solution after model optimization are compared. There are two sets of data in total, and the comparison results are shown in Table 2: Table 2 Comparison of historical data processing solutions

[0019] It can be seen from the optimized plan that the optimized plan is basically the same as the historical final plan, and the optimized plan gives a more reasonable plan. It also gives cost data and the cost data are lower than the historical initial plan.

[0020] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A water pipe intelligent processing method based on BIM, characterized in that: include: Calling the BIM model data of the building, and obtaining the water pipe line data and the water pipe laying processing plan according to the BIM model data; Acquire abnormal line points of the water pipe line data of the historical BIM model data, mark the historical water pipe line data according to the abnormal line points to generate first pipeline data; generate second pipeline data according to the water pipe line data; Establishing a water pipe anomaly matching model, matching the data anomaly features of the first pipeline data according to the second pipeline data, marking the second pipeline data according to the anomaly matching result, and generating third pipeline data and fourth pipeline data; Acquire a first target parameter according to the water pipe laying processing plan; extract the laying processing plan of the third pipeline data in the water pipe laying processing plan; call the abnormal laying processing plan of the corresponding first pipeline data according to the third pipeline data, and extract a first abnormal parameter according to the abnormal laying processing plan; Establishing a particle swarm optimization model, optimizing the laying and processing plan of the third pipeline data according to the first target parameter and the first abnormal parameter, and generating an abnormally optimized water pipe laying and processing plan; The water pipes of the third pipeline data are processed according to the abnormally optimized water pipe laying processing plan, and the fourth pipeline data are processed according to the water pipe laying processing plan.

2. According to the BIM-based water pipe intelligent processing method of claim 1, it is characterized in that: Acquire abnormal line points of the water pipe line data of the historical BIM model data, and mark the historical water pipe line data according to the abnormal line points to generate first pipeline data; Generating second pipeline data according to the water pipeline line data includes: Establishing a three-dimensional coordinate system according to the BIM model data of the input water pipeline line, wherein the input water pipeline line includes the water pipeline line data and the historical water pipeline line data; mapping the historical water pipeline line to the three-dimensional coordinate system to generate input water pipeline line coordinates; segmenting the input water pipeline line coordinates to generate an input water pipeline line coordinate set; According to the mapping of the three-dimensional coordinate system into a two-dimensional coordinate system, the coordinates of the input water pipe line coordinate set are projected into the two-dimensional coordinate system to generate an input water pipe line transposed coordinate set; according to the coordinates of the input water pipe line transposed coordinate set, a plurality of input water pipe line curves are generated; the input water pipe line curve includes the second pipeline data and the historical water pipe line; If the input water pipeline route is the historical water pipeline route, the coordinates of the abnormal route points marked by the historical water pipeline route coordinates are transposed to abnormal route transposition points, and multiple segments of the input water pipeline route curves are marked according to the abnormal route transposition points to generate the first pipeline data.

3. The water pipe intelligent processing method based on BIM according to claim 2 is characterized in that: The abnormal line transposition points include multiple groups of data, and every two abnormal line transposition points form a group of data.

4. The water pipe intelligent processing method based on BIM according to claim 1 is characterized in that: Establishing a water pipe anomaly matching model, matching the data anomaly features of the first pipeline data according to the second pipeline data, marking the second pipeline data according to the anomaly matching result, and generating third pipeline data and fourth pipeline data, including: The water pipe anomaly matching model includes a line data compression unit, a line feature extraction unit, a line feature classification unit and an abnormal line output unit; The line data compression unit performs data standardization and data encoding on the second pipeline data; The line feature extraction unit extracts data features of the standardized second pipeline data; the line feature extraction unit includes a residual convolution network, an LSTM network and a feature fusion layer, the residual convolution network performs data feature enhancement and morphological feature analysis on the second pipeline data, the LSTM network extracts trend features of the second pipeline data, and the feature fusion layer fuses data features of the residual convolution network and the LSTM network; The line feature classification unit performs feature classification on the second pipeline data according to the feature label of the first pipeline data, and matches the feature label of the first pipeline data according to the classification result; The abnormal line output unit generates and outputs a feature label of the second pipeline data.

5. The water pipe intelligent processing method based on BIM according to claim 4 is characterized in that: The number of the residual convolutional networks is 2, the number of the LSTM networks is 6, the feature fusion layer is Relu, and the line feature classification unit includes a fully connected layer and a Bayesian classifier.

6. The water pipe intelligent processing method based on BIM according to claim 1 is characterized in that: A particle swarm optimization model is established, and the laying processing plan of the third pipeline data is optimized according to the first target parameter and the target first abnormal parameter, and the abnormal optimized water pipe laying processing plan is generated, including: setting a priority weight for the first target parameter according to the water pipe laying processing plan, and initializing the speed and position of the particle swarm with the first target parameter; using the priority weight and the target first abnormal parameter as penalty coefficients, and using the parameters of the water pipe laying processing plan as target conditions, iterating the position and speed of the particle swarm for multiple times until the particle swarm reaches the parameters of the water pipe laying processing plan, and generating the abnormal optimized water pipe laying processing plan according to the first target parameter corresponding to the optimized particle swarm.

7. The water pipe intelligent processing method based on BIM according to claim 6 is characterized in that: The priority weights are cost weight and length weight; the first target parameters are the number of water pipes and the length of water pipes; the target first abnormal parameters include the number of abnormal water pipes and the length of abnormal water pipes; the parameters of the water pipe laying and processing plan are the total length of the third pipeline data that needs to be optimized.

8. A water pipe intelligent processing system based on BIM, characterized by: A water pipe line acquisition module is used to acquire BIM model data of a building, and acquire water pipe line data and a water pipe laying and processing plan based on the BIM model data; and generate second pipeline data based on the water pipe line data; An abnormal line acquisition module is used to acquire abnormal line points of the water pipe line data of the historical BIM model data, and to generate first pipeline data according to marking the historical water pipe line data with the abnormal line points; An abnormal line matching module is used to establish a water pipe abnormal matching model, match the data abnormality characteristics of the first pipeline data according to the second pipeline data, mark the second pipeline data according to the abnormal matching result, and generate third pipeline data and fourth pipeline data; The abnormal parameter extraction module obtains the first target parameter according to the water pipe laying processing plan; extracts the laying processing plan of the third pipeline data in the water pipe laying processing plan; calls the abnormal laying processing plan of the corresponding first pipeline data according to the third pipeline data, and extracts the first abnormal parameter according to the abnormal laying processing plan; The abnormal line optimization module establishes a particle swarm optimization model, optimizes the laying and processing plan of the third pipeline data according to the first target parameter and the first abnormal parameter, and generates an abnormal optimized water pipe laying and processing plan; The normal-abnormal line processing plan generating module processes the water pipes of the third pipeline data according to the abnormal optimized water pipe laying processing plan, and processes the fourth pipeline data according to the water pipe laying processing plan.

9. The BIM-based water pipe intelligent processing system according to claim 8 is characterized in that: The water pipe anomaly matching model includes a line data compression unit, a line feature extraction unit, a line feature classification unit and an abnormal line output unit. The line feature extraction unit includes 2 residual convolutional networks, 6 LSTM networks and 1 Relu feature fusion layer; the line feature classification unit includes a fully connected layer and a Bayesian classifier.

10. The BIM-based water pipe intelligent processing system according to claim 8, characterized in that: The abnormal line optimization module includes: setting a priority weight for the first target parameter based on the water pipe laying processing plan, and initializing the speed and position of the particle swarm with the first target parameter; using the priority weight and the target first abnormal parameter as a penalty coefficient, and using the parameters of the water pipe laying processing plan as target conditions, iterating the position and the speed of the particle swarm multiple times until the particle swarm reaches the parameters of the water pipe laying processing plan, and generating the abnormal optimized water pipe laying processing plan according to the first target parameter corresponding to the optimized particle swarm.

Images