Construction site intelligent arrangement system and method based on CAD drawings

Abstract

The application discloses a kind of construction site intelligent arrangement systems and methods based on CAD drawing, belong to civil engineering intelligent construction field.The arrangement system includes drawing information acquisition module and generating network combination module;Drawing information acquisition module can be converted into drawing image by the construction plan to be arranged;Generating network combination module includes construction site arrangement generation module, first discrimination module and second discrimination module;First discrimination module and second discrimination module can train construction site arrangement generation module;The construction site arrangement generation module of training completion can output construction plan arrangement drawing after the drawing image of the construction plan to be arranged is input.The arrangement system can directly generate construction plan layout based on the drawing image of the construction plan to be arranged, and construction site arrangement generation module is trained by first discrimination module and second discrimination module together, with the advantages of fast arrangement speed and good arrangement effect.

Description

Technical Field

[0001] This invention relates to an intelligent construction site layout system and method based on CAD drawings, belonging to the field of intelligent construction in civil engineering. Background Technology

[0002] In the process management of building structure construction, the layout of the construction site needs to be adjusted according to the construction progress, which often involves multiple changes. The process of drawing a construction site layout plan is to determine the location of construction areas, temporary facilities, construction sites, construction equipment, etc., based on the original plan of the construction site, taking into account factors such as safety, external environmental factors, traffic routes, temporary facilities, and material storage, and then express them on the drawings for use in on-site construction.

[0003] Currently, the layout of construction site plans mainly relies on the experience and intuition of construction workers. This process is time-consuming and labor-intensive, and may not even achieve the optimal layout effect. Summary of the Invention

[0004] To address the problems of time-consuming, labor-intensive, and difficult-to-achieve optimal layout in existing construction site plan layouts, this invention proposes an intelligent construction site layout system and method based on CAD drawings, which can automatically generate a construction site layout plan based on the construction plan to be laid out.

[0005] To solve the above technical problems, the present invention includes the following technical solutions:

[0006] A construction site intelligent layout system based on CAD drawings includes a drawing information acquisition module and a network combination generation module;

[0007] The drawing information acquisition module can convert the construction plan to be arranged into drawing images;

[0008] The generative network assembly module includes a construction site layout generation module, a first discrimination module, and a second discrimination module. The first discrimination module and the second discrimination module can train the construction site layout generation module. After training, the construction site layout generation module can output a construction site layout diagram after inputting a drawing image of the construction plan to be arranged.

[0009] The first discrimination module, based on machine learning technology, can determine whether the construction site layout plan is correct; the second discrimination module can use large model technology to learn the relevant specifications for construction site layout and the construction experience of project personnel, and determine whether the construction layout plan is correct by collecting text data of the construction site layout plan.

[0010] The training process of the construction site layout generation module is as follows: the first and second discrimination modules judge the rationality of the construction plan layout output by the construction site layout generation module, and feed the judgment results back to the construction site layout generation module. The construction site layout generation module adjusts the internal network weights. The construction site layout generation module is trained in multiple rounds. The training is completed when the loss function is less than a certain threshold or the number of training rounds exceeds the preset number.

[0011] This invention also provides an intelligent construction site layout method based on CAD drawings, which uses the aforementioned intelligent construction site layout system based on CAD drawings for layout. The layout method includes the following steps:

[0012] Step 1: Convert the construction plan to be arranged into a drawing image using the drawing information acquisition module;

[0013] Step 2: Input the drawing image of the construction plan to be arranged into the input terminal of the construction site layout generation module, and the construction site layout generation module outputs the construction plan layout.

[0014] The construction site layout generation module is trained by a first discrimination module and a second discrimination module. The training process is as follows: the input terminal of the construction site layout generation module is the construction plan to be arranged, and the output is the construction plan layout for training; the first discrimination module determines whether the construction plan layout is correct based on machine learning technology; the second discrimination module has learned the relevant specifications and construction experience of project personnel for construction site layout, and determines whether the construction plan layout is correct by collecting text data of the construction plan layout; the construction site layout generation module adjusts its internal network weights according to the judgment results of the feedback from the first and second discrimination modules; the construction site layout generation module is trained in multiple rounds, and the training is completed when the loss function is less than a certain threshold or the number of training rounds exceeds a preset number.

[0015] Furthermore, several construction site layout diagrams labeled as reasonable and those labeled as unreasonable are pre-set, and the first discrimination module is trained. After training, the first discrimination module can identify whether the construction site layout diagram is correct.

[0016] Furthermore, the construction site layout diagram marked with reasonable labels is obtained by collecting historical data, converting the construction site layout diagram into a drawing image, and then marking it with reasonable labels.

[0017] The construction site layout diagram marked as unreasonable is obtained by collecting historical data, making unreasonable modifications to some of the site types in the construction site layout diagram, and then converting it into a drawing image and marking it as unreasonable.

[0018] Furthermore, the types of sites arranged in the construction site layout diagram output by the construction site layout generation module are distinguished by preset colors.

[0019] During the training process of the construction site layout generation module, the second discrimination module can identify the site type, location and size based on color, and convert the site type, location and size into text data.

[0020] The present invention, by adopting the above technical solutions, has the following advantages and positive effects compared with the prior art: The intelligent construction site layout system and method based on CAD drawings provided by the present invention can convert the construction plan to be arranged from CAD drawings into drawing images through the drawing information acquisition module, and then directly generate the construction plan layout drawing based on the drawing image of the construction plan to be arranged. It has the advantage of fast layout speed and avoids rework and repeated modifications in manual layout. Moreover, the construction site layout generation module is trained by the first discrimination module and the second discrimination module together, and the generated construction plan layout has good layout effect. Attached Figure Description

[0021] Figure 1 This is an image drawing of the construction plan to be arranged in this invention;

[0022] Figure 2 This is a construction plan layout diagram for this invention;

[0023] Figure 3 This is a flowchart of the training process for the construction site layout generation module in this invention.

[0024] Figure 4 This is a flowchart of the intelligent construction site layout method based on CAD drawings in this invention. Detailed Implementation

[0025] The intelligent construction site layout system and method based on CAD drawings provided by the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. The advantages and features of the present invention will become clearer with the following description. It should be noted that the accompanying drawings are all in a very simplified form and use non-precise scales, and are only used to facilitate and clarify the illustration of the embodiments of the present invention.

[0026] Example 1

[0027] Combination Figure 1 and Figure 2As shown, this embodiment provides an intelligent construction site layout system based on CAD drawings, used to automatically draw construction site layout diagrams. Drawing a construction site layout diagram involves determining the locations of various site types, such as entrances, construction roads, temporary water and electricity areas, temporary vehicle areas, rebar processing areas, office areas, and vertical transportation areas, based on the original site plan and considering factors such as safety, external environmental factors, traffic routes, temporary facilities, and material storage. These locations are then represented on the drawings to guide on-site construction. The original site plan is the construction site plan to be laid out, which typically includes the outline of the construction area and the floor plan of the buildings.

[0028] The layout system includes a drawing information acquisition module and a generation network assembly module. The drawing information acquisition module converts the construction site layout plan to be arranged into a drawing image. For example, converting the construction site layout plan into a drawing image can be achieved using methods such as, but not limited to, CAD printing and redrawing of graphic elements. The generation network assembly module includes a construction site layout generation module, a first discrimination module, and a second discrimination module. The first and second discrimination modules train the construction site layout generation module. Once trained, the construction site layout generation module can output a construction site layout plan that meets the requirements after inputting a drawing image of the construction site layout plan to be arranged. In other words, the generation network assembly module includes a training phase and a usage phase. During the training phase, the first and second discrimination modules are used to train the construction site layout generation module. After training is complete, the module is used in the usage phase to directly generate a construction site layout plan that meets the requirements.

[0029] Before training the construction site layout generation module using the first and second discriminant modules, both modules must be trained first to meet certain requirements before they can be used to train the construction site layout generation module. For example, the first discriminant module may include, but is not limited to, a visual model including a convolutional neural network, while the second discriminant module uses a fine-tuned model trained from a large-model discriminant module.

[0030] To train the first discrimination module, training data needs to be set up, including several construction site layouts labeled as "reasonable" and several labeled as "unreasonable." The first discrimination module will then be able to identify whether a construction site layout is correct and provide a judgment of whether it is reasonable or unreasonable. Construction site layouts labeled as "reasonable" can be obtained by collecting historical data, converting the construction site layout into a drawing image, and then labeling it as "reasonable." Construction site layouts labeled as "unreasonable" can be obtained by collecting historical data, making unreasonable modifications to some site types in the construction site layout, converting it into a drawing image, and then labeling it as "unreasonable."

[0031] The second discrimination module can employ a large-model discrimination module. This module utilizes large-model technology to learn from existing construction site layout specifications and project personnel's construction experience, performing pre-training and fine-tuning. Once trained, the second discrimination module can identify text data in the construction site layout diagram output by the construction site layout generation module. Based on the learned layout specifications and construction personnel experience, it determines whether the construction layout diagram meets construction requirements and feeds the judgment result back to the construction site layout generation module.

[0032] In one specific embodiment, the site types in the construction site layout diagram output by the construction site layout generation module are distinguished by preset colors. The second discrimination module can identify the corresponding layout area based on the color and convert it into text data. The recognition methods include, but are not limited to, visual models using convolutional neural networks. The text data includes the site type, location, and size of the layout area. The site types include entrances, construction roads, temporary water and electricity areas, temporary vehicle areas, rebar processing areas, office areas, vertical transportation areas, etc. The location and size of the layout areas can be determined based on visual recognition technology. For example, in a construction site layout diagram, the outer contour of the site is represented in red, the rebar processing area in green, the structure to be built in black, the entrance is represented by a red arrow, and the area to be demolished is represented by a gray shading, etc.

[0033] Example 2

[0034] This embodiment provides a method for intelligent layout of construction sites based on CAD drawings. The method uses the intelligent layout system for construction sites based on CAD drawings from Embodiment 1. The layout method includes the following steps:

[0035] Step 1: Convert the construction plan to be arranged into a drawing image using the drawing information acquisition module;

[0036] Step 2: Input the drawing image of the construction plan to be arranged into the input terminal of the construction site layout generation module, and output the construction plan layout diagram.

[0037] The construction site layout generation module is trained by a first discrimination module and a second discrimination module. The training process is as follows: the input terminal of the construction site layout generation module is the construction plan to be arranged, and the output is the construction plan layout for training; the first discrimination module determines whether the construction plan layout is correct based on machine learning technology; the second discrimination module has learned the relevant specifications and construction experience of project personnel for construction site layout, and determines whether the construction plan layout is correct by collecting text data of the construction plan layout; the construction site layout generation module adjusts its internal network weights according to the judgment results of the feedback from the first and second discrimination modules; the construction site layout generation module is trained in multiple rounds, and the training is completed when the loss function is less than a certain threshold or the number of training rounds exceeds a preset number.

[0038] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0039] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the appended claims.

Claims

1. A construction site intelligent layout system based on CAD drawings, characterized in that, The layout system includes a drawing information acquisition module and a network combination generation module; The drawing information acquisition module can convert the construction plan to be arranged into drawing images; The generative network assembly module includes a construction site layout generation module, a first discrimination module, and a second discrimination module; the first discrimination module and the second discrimination module can train the construction site layout generation module; The trained construction site layout generation module can output a construction site layout diagram after inputting a drawing image of the construction plan to be arranged. The first discrimination module, based on machine learning technology, can determine whether the construction site layout plan is correct; the second discrimination module can use large model technology to learn the relevant specifications for construction site layout and the construction experience of project personnel, and determine whether the construction layout plan is correct by collecting text data of the construction site layout plan. The training process of the construction site layout generation module is as follows: the first and second discrimination modules judge the rationality of the construction plan layout output by the construction site layout generation module, and feed the judgment results back to the construction site layout generation module. The construction site layout generation module adjusts the internal network weights. The construction site layout generation module is trained in multiple rounds. The training is completed when the loss function is less than a certain threshold or the number of training rounds exceeds the preset number. The types of sites arranged in the construction site layout diagram output by the construction site layout generation module are distinguished by preset colors. During the training process of the construction site layout generation module, the second discrimination module can identify the types, locations and sizes of the arranged sites based on the colors, and convert the types, locations and sizes of the arranged sites into text data.

2. A method for intelligent layout of construction sites based on CAD drawings, characterized in that, The construction site layout system based on CAD drawings, as described in claim 1, is used for layout, and the layout method includes the following steps: Step 1: Convert the construction plan to be arranged into a drawing image using the drawing information acquisition module; Step 2: Input the drawing image of the construction plan to be arranged into the input terminal of the construction site layout generation module, and the construction site layout generation module outputs the construction plan layout. The construction site layout generation module is trained by a first discrimination module and a second discrimination module. The training process is as follows: the input terminal of the construction site layout generation module is the construction plan to be arranged, and the output is the construction plan layout for training; the first discrimination module determines whether the construction plan layout is correct based on machine learning technology; the second discrimination module has learned the relevant specifications and construction experience of project personnel for construction site layout, and determines whether the construction plan layout is correct by collecting text data of the construction plan layout; the construction site layout generation module adjusts its internal network weights according to the judgment results of the feedback from the first and second discrimination modules; the construction site layout generation module is trained in multiple rounds, and the training is completed when the loss function is less than a certain threshold or the number of training rounds exceeds a preset number.

3. The intelligent construction site layout method based on CAD drawings as described in claim 2, characterized in that, Several construction site layout diagrams labeled as reasonable and those labeled as unreasonable are pre-set. The first discrimination module is then trained, and the trained first discrimination module can identify whether the construction site layout diagram is correct.

4. The intelligent construction site layout method based on CAD drawings as described in claim 3, characterized in that, The construction site layout plan marked with appropriate labels is obtained by collecting historical data, converting the construction site layout plan into a drawing image, and then marking it with appropriate labels. The construction site layout diagram marked as unreasonable is obtained by collecting historical data, making unreasonable modifications to some of the site types in the construction site layout diagram, and then converting it into a drawing image and marking it as unreasonable.

Images