CN114626294A - Planning layout hierarchical generation method for university campus - Google Patents
Planning layout hierarchical generation method for university campus Download PDFInfo
- Publication number
- CN114626294A CN114626294A CN202210181789.XA CN202210181789A CN114626294A CN 114626294 A CN114626294 A CN 114626294A CN 202210181789 A CN202210181789 A CN 202210181789A CN 114626294 A CN114626294 A CN 114626294A
- Authority
- CN
- China
- Prior art keywords
- model
- layout
- planning
- campus
- university
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/10—Geometric CAD
- G06F30/13—Architectural design, e.g. computer-aided architectural design [CAAD] related to design of buildings, bridges, landscapes, production plants or roads
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Geometry (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- Software Systems (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Medical Informatics (AREA)
- Mathematical Physics (AREA)
- Architecture (AREA)
- Civil Engineering (AREA)
- Data Mining & Analysis (AREA)
- Computing Systems (AREA)
- Structural Engineering (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention relates to a hierarchical generation method for university campus planning layout, which generates a confrontation network model based on pix2pix, provides a series of unconventional working methods in aspects of case selection, processing annotation and the like, and innovatively provides a hierarchical learning method, and comprises the following steps: collecting and screening university campus planning plane layout cases with the same spatial characteristics; unifying standard marking cases; training by a machine learning algorithm; and (4) planning and layout hierarchical generation of the university campus. The university campus planning design is integrated into a computer algorithm, a learning framework based on a machine learning algorithm is constructed, a series of specific case screening rules, redrawing rules and marking rules are provided mainly aiming at the intrinsic rules of the university campus and the relation between the intrinsic rules and a city, the learning effect of the machine learning algorithm on the university campus planning design layout scheme is improved through a specific marking method, a planning designer is assisted in designing, and the design efficiency is improved.
Description
Technical Field
The invention belongs to the technical field of planning design generation, and particularly relates to a planning layout hierarchical generation method for a university campus.
Background
The appearance of the artificial intelligence algorithm technology brings a new breakthrough to the traditional industries such as building design and the like, and in the aspect of building design practice, the house type samples are artificially screened and labeled, so that the machine can effectively learn the plane layout of the building apartment, and the strong potential of the deep learning technology in the aspect of building plane layout generation is expressed. However, the machine learning technology at present has not been mature for learning the planning plane layout with larger scale, and there is no unified and effective case screening rule and marking rule for the planning plane layout with large scale.
In the field of planning and design, a large amount of time cost, manpower and material resources are consumed at the stage of depending on manpower to design a scheme at present, the quality of the generated scheme is uneven, the quality depends on the level and experience of a designer to the greatest extent, and the effect of scheme generation in the later stage cannot be estimated quickly in the early stage of scheme design. Meanwhile, campus planning cases have limited access, are difficult to form a huge sample library and generally show diversified characteristics due to different design styles of designers, and if the cases are not classified, screened and labeled and are all input into a computer, the generated result is not ideal (Linwen is strong, and the design layout of a primary school based on deep learning automatically generates research [ D ]. the university of south China's Articians, 2020.).
Disclosure of Invention
In combination with the above problems, the invention provides a method for generating a confrontation network model based on artificial intelligence pix2pix, and aiming at specific spatial characteristics, the method for generating a planning layout of a university campus in a grading manner. The core of this method is: a set of complete work flows of case selection, case labeling and hierarchical learning is provided on the basis of campus design planning experience, and the set of work flows are specifically provided on the basis of the generated target and are completely different from conventional artificial intelligence learning. In order to assist a designer to design in the early stage of scheme design, quickly generate a design scheme and save the time cost of repeated modification, the invention uses the existing pix2pix to generate a confrontation network model, and provides a set of complete work flows of case selection, case annotation and hierarchical learning on the basis of the campus design planning experience, wherein the work flows are specifically provided on the basis of the generated target and are completely different from the conventional artificial intelligence learning, and finally the generation of a planning design layout scheme of a university campus with specific spatial characteristics is realized.
The invention is realized by at least one of the following technical schemes.
A planning layout hierarchical generation method for a university campus comprises the following steps:
s1, collecting planning and designing plane layout cases of the university campus as an original data set;
s2, marking the original data set and expanding data based on a unified standard principle;
s3, training the labeled data set based on a machine learning algorithm to obtain a machine learning model, and testing the machine learning model;
s4, the university campus planning layout is generated in a grading way: the method comprises the steps of dividing campus internal functions, generating campus internal function partitions according to land conditions, generating building function layouts according to the campus internal function partitions, and finally achieving automatic generation of planning design layout schemes of university campuses.
Preferably, in step S1, planning and design plane layout cases of the university campus with the same characteristics are collected, and the types of the campus planning and design planes are filtered by the architect using the related building knowledge.
Preferably, the marked content comprises a preset drawing size, a preset drawing proportion and an image of each functional building monomer standard prototype, and the image of the original data set is adjusted to be a uniform size according to the preset drawing size and the preset drawing proportion.
Preferably, in step S2, the data is augmented by data enhancement, the data amount is increased by horizontally and vertically flipping the image, and the data is divided into a training image set and a test image set.
Preferably, the training labeled data set comprises:
presetting a machine learning model and machine learning model parameters, adjusting the machine learning model parameters, and training and verifying a training image set and a test image set to obtain set parameters;
the machine learning model generates a confrontational network model or pix2pix HD model for pix2 pix.
Preferably, the pix2pix generation countermeasure network model is used to handle a pair of image translation problems, which is a process of obtaining a desired output image based on an input image, and can be regarded as a mapping between images.
Preferably, the pix2pix HD model is optimized on the pix2pix model generation countermeasure network model, generating semantic edits to high resolution images and pictures.
Preferably, the step of hierarchical training for pix2pix to generate the antagonistic network model or the pix2pixHD model is as follows:
inputting the surrounding field roads, the functions of the surrounding blocks, and the functional partitions and the road network in the university as paired data into a pix2 pix-generated confrontation network model or a pix2pixHD model for learning to obtain a first model;
inputting the functional partitions and the road network in the university and the layout plan of the real university campus as paired data into a pix2pix model to generate a confrontation network model or a pix2pixHD model for learning to obtain a second model;
and (3) testing a model: and inputting the test image set into the first model and the second model respectively for the trained models, and generating images.
Inputting the functions of the roads of the surrounding field and the surrounding blocks into a first model, and generating functional partitions and a road network in a university;
and inputting the generated functional partitions and the road network inside the university into a second model to generate a final university campus layout plan so as to obtain a university campus layout result which is mainly characterized by forming a central area by a loop surrounding landscape, and continuously adjusting the parameters of the machine learning model until a reasonable generation result is obtained.
Preferably, the land conditions comprise city arterial roads, water bodies, mountain bodies and blocks around the field.
Preferably, the planning layout scheme of the university campus is finally generated and exported to a file of a preset type, wherein the preset type comprises a bitmap and a vector diagram.
Compared with the prior art, the invention has the beneficial effects that:
the invention integrates planning design into a computer algorithm, constructs a layout learning framework based on a machine learning algorithm, generates an confrontation network model based on the existing pix2pix, forms innovation points in aspects of case selection, processing, labeling and the like, provides a series of effective unconventional working methods, innovatively provides a working process of staged learning, improves the learning effect of the machine learning algorithm on large-scale land, assists a planning designer to design, greatly shortens the time, improves the design efficiency, realizes the automatic generation of a layout scheme for large-scale planning land, and can be based on the original field conditions: site boundaries, surrounding roads, surrounding blocks and the like, a feasible university campus layout scheme is quickly generated, and a design idea with more possibilities is provided.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention.
FIG. 1 is a flowchart of an embodiment of a method for hierarchical generation of a planned layout for a university campus;
FIG. 2 is an embodiment data collection and expansion flow diagram;
figure 3 is a flow diagram of an embodiment of generating a university campus planning layout.
Detailed Description
The invention is described in detail below with reference to the figures and specific embodiments. The present embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation manner and a specific operation process are given, but the scope of the present invention is not limited to the following embodiments.
Example 1
A planning layout hierarchical generation method for a university campus comprises the following steps:
s1, select and collect cases: collecting and screening planning and design plan layout cases of a university campus with a specific space or the same spatial characteristics to obtain an original data set; unified labeling, namely labeling the original data set based on a unified standard labeling method; training a labeling data set by an artificial intelligence algorithm to obtain an algorithm model, and testing the model;
aiming at an internal operation rule of a university campus and the relation between the internal operation rule and a city, a planning layout case of the university campus is collected and screened as an original data set according to the same spatial characteristics, and the steps are as follows:
1.1) data collection: collecting pictures of planning and designing examples of a university campus in related data;
1.2) case screening: and screening the collected pictures according to the spatial characteristics of the grid university planning layout to obtain cases with the same spatial characteristics, wherein the cases have similar plane characteristics and logics, such as screening out a central loop type campus layout or a grid type campus layout.
2) The method is characterized in that unified labeling is carried out on the internal operation rule of the university campus and the relation between the internal operation rule and the city, and the method comprises the following steps based on a unified standard labeling case:
2.1) uniformly proportioning the graph paper;
2.2) marking layout elements by color blocks, and simplifying the total plan map into color block images;
2.3) building a standard prototype image of the building monomer;
3) the artificial intelligence algorithm carries out image training and comprises the following steps:
3.1) carrying out planning layout classification training and testing on a university campus, firstly generating an urban road network and a planning functional block according to land use conditions, then generating functional partitions inside the block, and finally generating functional partition internal building function layout;
the right-of-way condition is determined according to at least one of the following information: urban arterial road, water body, mountain, neighborhood around the place, etc.
And S2, generating the university campus planning layout in a grading manner. Firstly, performing campus internal function division, and generating a campus internal function partition according to a land use condition; and generating a building layout, and generating a building function layout according to the campus internal function partitions. And finally, automatically generating a planning design layout scheme of the university campus.
The planning layout of the university campus is generated in a grading manner, and the steps are as follows:
4.1) inputting the land conditions to generate the internal function partitions of the university campus;
and 4.3) inputting the function partitions in the university campus to generate a building function layout.
Example 2
As shown in fig. 1, a method for generating a planning layout hierarchy of a university campus specifically includes the following steps:
step 1, collecting and screening large-scale land planning layout cases with the same regularity, and facilitating computer discovery of implicit laws therein, the method comprises the following specific steps:
step 1.1, data collection: pictures of planning and designing examples are collected in works sets of various major design houses, journal papers, conference papers and academic papers related to research and planning, and related books of planning and designing.
Step 1.2, case screening: the collected pictures are screened according to a spatial mode with the same fishbone layout characteristics.
And 2, marking cases based on unified standards. The method comprises the following steps:
step 2.1, size and scale unification, in the collected cases, the total plan with scale is unified by scale, the plan without scale is quasi-unified by the same building blocks in the site, and the unified plan is centered on a blank sheet with size of 300mm x 300mm, pixels of 3543 x 3543, and resolution of 300 pixels/inch.
And 2.2, marking layout elements by color blocks, simplifying the general plan view, and representing important layout elements by different color blocks.
And 2.3, establishing a standard prototype of the building monomer. The building is marked by blocks only, and the concave-convex details of the shapes such as staircases and the like are ignored. The building in each functional area is subjected to standardized design, a building monomer standard prototype is given, and the building monomer is allowed to be subjected to limited change on the basis of the prototype when the building monomer is drawn, such as lengthening or shortening the long side of the building.
And 2.4, expanding data. Because the number of the original collected data sets is small, the data sets are expanded in a data enhancement mode, and the data volume is increased by horizontally turning and vertically turning the original campus layout image.
And 2.5, data arrangement. And classifying the obtained data into a training image data set and a test image data set, wherein 4/5 of the preset total data is used as the training image set, and 1/5 of the preset total data is used as the test image set.
And 3, carrying out image training and testing by an artificial intelligence algorithm. The method comprises the following steps:
step 3.1, selecting an artificial intelligence algorithm model and adjusting parameters, wherein the artificial intelligence algorithm model is determined according to one of the following information: pix2pix generates a confrontation network model, a pix2pix HD model, etc.; according to the definition requirement of the generated image, if the pixel requirement of the image is high, a pix2pix HD model is used; if the image pixel requirements are low, pix2pix is used to generate the confrontation network model. And training and verifying the training image set and the test image set for multiple times to obtain a better setting parameter.
3.2, generating a campus internal function partition according to the land conditions, training through a machine learning algorithm, and establishing a first machine learning model; and then generating a building function layout according to the campus internal function partitions, training through a machine learning algorithm, and establishing a second machine learning model.
And 3.3, testing the trained first machine learning model and the trained second machine learning model respectively by using a test set. And respectively inputting the prepared test image sets into the two machine learning models, generating images, comparing and analyzing the images with the original campus layout image, and verifying the learning effectiveness.
And 4, generating an image by machine learning. The method comprises the following specific steps:
step 4.1, inputting a land use condition to generate a campus internal function partition;
and 4.3, inputting the internal function partitions of the campus to generate a building function layout.
Example 3
The invention will now be illustrated by means of an example, which is particularly applicable to the automatic generation of a university campus planning design layout, mainly characterized by a central loop type campus level layout:
(I) case data Collection and screening
Step 1, data collection: pictures of university campus design examples are collected from works sets of various university schools, journal papers, conference papers and academic position papers related to the university campus planning and related books of the university campus planning and design, and 30 data are obtained as a basic data set.
Step 1.2, case screening: aiming at the internal operation rule of the university campus and the relation between the internal operation rule and the city, the collected pictures are screened according to the same spatial mode, the method selects the central loop type campus plane layout characteristic as an example for explanation, and mainly follows the following rules: the natural conditions in the land are simple, the terrain is generally flat, a main water body is contained, and a small number of mountain bodies can be distributed; selecting a case which is subjected to spatial organization by taking a central area formed by a loop around a landscape as a main characteristic; the land utilization scale of the university is unified by about 50 to 100 hectares; fourthly, the field shape is relatively square.
(II) unified processing of raw data information
And 2.1, unifying the proportion, wherein in the collected cases, the total plan with the scale is unified by the scale, the plan without the scale is quasi-unified by a 400m standard track and field in the field, and the drawing with the unified proportion is centrally placed on a blank paper with the size of 300 mm/300 mm, the pixel of 3543/3543 and the resolution of 300 pixels/inch.
And 2.2, marking layout elements by color blocks, simplifying the general plan view, and representing 9 important layout elements by different color blocks.
And 2.3, establishing a standard prototype of the building monomer. The building is marked by squares only, and the concave-convex details of the shapes of staircases and the like are ignored. The building in each functional area is subjected to standardized design, a building monomer standard prototype is given, and the building monomer is allowed to be subjected to limited change on the basis of the prototype when the building monomer standard prototype is manually drawn through photoshop software, such as the long side of the building is prolonged or shortened.
And 2.4, expanding data. Because the number of the original collected data sets is small, the data is expanded in a data enhancement mode, and the data volume is increased by performing two modes of horizontal turning and vertical turning on the original campus layout image, so that 120 pairs of data are obtained.
And 2.5, data arrangement. 100 pairs are used as training image sets, and 20 pairs are used as test image sets.
(III) hierarchical training of artificial intelligence algorithm
And 3.1, selecting a machine learning model and adjusting parameters, selecting and using pix2pix to generate a confrontation network model, and performing multiple training and verification on the training image set and the test image set to obtain a better setting parameter. And in the learning process, the parameters are tried and adjusted for multiple times, and finally, the result is found to be better when the learning rate is set to be 0.002 and the iteration number is about 2000.
Step 3.2, training pix2pix to generate a confrontation network model, and adopting a step-by-step training mode on the basis of the step 3.1, wherein the method specifically comprises the following steps:
inputting a surrounding site condition graph (surrounding site roads and surrounding block functions) and a real internal function partition graph (functional partitions and road networks in universities) as paired data into a pix2pix generation confrontation network model for learning, and establishing a university campus layout to generate a confrontation network first model;
inputting a real internal function partition graph (function partitions and a road network inside a university) and a real university campus layout plan as paired data into a pix2pix generation confrontation network model for learning, and establishing a university campus layout to generate a confrontation network second model;
and 3.3, testing the model, namely testing the trained model by using a test set. And respectively inputting the prepared test image sets into a university campus layout generation countermeasure network first model and a university campus layout countermeasure network second model, performing image generation, comparing and analyzing with the original campus layout image, and verifying the learning effectiveness.
Inputting a surrounding site condition graph (surrounding site roads and surrounding block functions) into a university campus layout to generate a first confrontation network model, and generating an internal function partition graph (function partitions and road networks in the university);
inputting the generated internal function partition map (function partitions and road networks inside the university) into the university campus layout to generate a second confrontation network model, and generating a final university campus layout plan map so as to obtain a university campus layout result with a central area formed by a loop around the landscape as a main characteristic;
and comparing the original scheme, and continuously adjusting the parameters of the machine learning model until a reasonable generation result is obtained.
(IV) realizing automatic layout generation
Inputting the new site condition diagram into the university campus layout to generate a first confrontation network model, generating a university function partitioning scheme, inputting the university function partitioning scheme into the university campus layout to generate a second confrontation network model, and automatically generating a university campus layout result with a central area formed by a loop surrounding landscape as a main characteristic.
The method can be applied to the layout generation of the university campus characterized by the grid roads, and the planning layout generation of the university campus characterized by the grid roads can be realized by changing the layout case of the university campus characterized by the grid roads in the step 1.
The method can be applied to the layout generation of the university campus characterized by the fishbone line type roads, and the planning layout generation of the university campus characterized by the fishbone line type can be realized by changing the university campus layout case characterized by the fishbone line type in the step 1.
The invention can automatically generate the layout of the university campus through the pix2pixHD model, and can realize the layout generation image of the university campus with higher definition images by changing the selection of the machine learning model in the step 3.1.
The foregoing is only illustrative of the embodiments of the present invention and application of the technical principles. The present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, although the present invention has been described by the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and is not limited to the embodiments.
Claims (10)
1. A planning layout hierarchical generation method for a university campus is characterized by comprising the following steps:
s1, collecting planning and designing plane layout cases of the university campus as an original data set;
s2, marking the original data set and expanding the data based on a unified standard principle;
s3, training the labeled data set based on the machine learning algorithm to obtain a machine learning model, and testing the machine learning model;
s4, the university campus planning layout is generated in a grading way: the method comprises the steps of dividing campus internal functions, generating campus internal function partitions according to land conditions, generating building function layouts according to the campus internal function partitions, and finally achieving automatic generation of a planning design layout scheme of a university campus.
2. The method as claimed in claim 1, wherein in step S1, the planning and design plan layout cases of the university campus with the same characteristics are collected, and the architect uses the related building knowledge to screen the types of the planning and design plans of the campus.
3. The method as claimed in claim 1, wherein the annotation content comprises a predetermined drawing size, a predetermined drawing proportion, and an image of a standard prototype of each functional building unit, and the image of the original data set is adjusted to a uniform size according to the predetermined drawing size and the predetermined drawing proportion.
4. The method as claimed in claim 1, wherein in step S2, the data is augmented by data enhancement, the data size is increased by flipping the image horizontally and flipping it vertically, and the data is divided into a training image set and a testing image set.
5. The method of claim 1, wherein training the annotated data set comprises:
presetting a machine learning model and machine learning model parameters, adjusting the machine learning model parameters, and training and verifying a training image set and a test image set to obtain set parameters;
the machine learning model generates a confrontation network model or pix2pix HD model for pix2 pix.
6. The method as claimed in claim 5, wherein the pix2pix generation countermeasure network model is used to deal with paired image translation problem, i.e. mapping between images.
7. The method of claim 5, wherein the pix2pix HD model is optimized on a pix2pix model generation countermeasure network model to generate semantic edits to images and pictures.
8. The method of claim 5, wherein the step of training pix2pix to generate the confrontation network model or pix2pixHD model in a hierarchical manner comprises:
inputting the functions of surrounding site roads and surrounding blocks and the function partitions and the road network in the university as paired data into a pix2pix generation countermeasure network model or a pix2pixHD model for learning to obtain a first model;
inputting the functional partitions and the road network in the university and the layout plan of the real university campus as paired data into a pix2pix model to generate a confrontation network model or a pix2pixHD model for learning to obtain a second model;
and (3) testing a model: respectively inputting the test image set into the first model and the second model for the trained model to generate images;
inputting the functions of the roads of the surrounding field and the surrounding blocks into a first model, and generating functional partitions and a road network in a university;
and inputting the generated functional partitions and the road network inside the university into a second model to generate a final university campus layout plan so as to obtain a university campus layout result which is mainly characterized by forming a central area by a loop surrounding landscape, and continuously adjusting the parameters of the machine learning model until a reasonable generation result is obtained.
9. The method as claimed in claim 1, wherein the land conditions include city arterial street, water body, mountain, and neighborhood of the site.
10. The method as claimed in any one of claims 1 to 9, wherein the final generated plan layout plan for the university campus is exported as a file of a preset type, and the preset type includes bitmap and vector graphics.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210181789.XA CN114626294A (en) | 2022-02-25 | 2022-02-25 | Planning layout hierarchical generation method for university campus |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210181789.XA CN114626294A (en) | 2022-02-25 | 2022-02-25 | Planning layout hierarchical generation method for university campus |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114626294A true CN114626294A (en) | 2022-06-14 |
Family
ID=81900173
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210181789.XA Pending CN114626294A (en) | 2022-02-25 | 2022-02-25 | Planning layout hierarchical generation method for university campus |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114626294A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117235854A (en) * | 2023-09-15 | 2023-12-15 | 东南大学建筑设计研究院有限公司 | Digital generation method for quantitative shape overall in university campus planning and design |
-
2022
- 2022-02-25 CN CN202210181789.XA patent/CN114626294A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117235854A (en) * | 2023-09-15 | 2023-12-15 | 东南大学建筑设计研究院有限公司 | Digital generation method for quantitative shape overall in university campus planning and design |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Li et al. | Modelling sustainable urban development by the integration of constrained cellular automata and GIS | |
| WO2022047958A1 (en) | Artificial intelligence-based urban road network automatic generation method | |
| CN106599332A (en) | Three-dimensional digital scheme aided design and display method | |
| CN112100718B (en) | Urban building group design method based on intelligent algorithm and energy consumption comprehensive optimization | |
| CN114092697B (en) | Building facade semantic segmentation method with attention fused with global and local depth features | |
| CN115470707A (en) | City scene simulation system | |
| Kaya et al. | Urban DNA: Morphogenetic analysis of urban pattern | |
| CN113963104B (en) | Three-dimensional model generation method of layout image generated based on GAN model | |
| CN114626294A (en) | Planning layout hierarchical generation method for university campus | |
| Turkienicz et al. | CityZoom: a visualization tool for the assessment of planning regulations | |
| CN116561845B (en) | Digital technology energized urban roof greening multi-scenario simulation method and system | |
| Bai | Sustainable Development Garden Landscape Design Based on Data Mining and Virtual Reality | |
| Yang et al. | Construction and Application of a Landscape Design Teaching Platform Driven by Artificial Intelligence | |
| Chahbi | Application Of Genetic Algorithm Classification Approach to Study Urban Streets Morphology at Neighborhood Scale | |
| Zhang et al. | An energy-saving design method for residential building group based on convolutional neural network | |
| KIRAZ et al. | Integrating user-behaviour as performance criteria in conceptual parametric design | |
| Pitts et al. | A parametric approach to 3D massing and density modelling | |
| Doumpioti et al. | Field Conditions | |
| Liu et al. | Graph Constrained Multiple Schemes Generation for Campus Layout | |
| Akodad et al. | Computational Approaches for Urban Design Within the MENA Region | |
| Fan et al. | Analysis and Reconstruction Method of Spatial Characteristics of Traditional Chinese Villages Based on Parameterization | |
| Niu et al. | A novel optimal road layout model based on urban cloud data | |
| Cha et al. | Developing algorithmic methodology to visualize & Evaluate the dynamics of urban form in the early design phase | |
| Ye et al. | Autonomous driving road network generation and evaluation based on generative artificial intelligence | |
| Mohseni Fard Naghani et al. | Ecological Reclamation of Distressed Urban Fabric through Open and Green Space Networks to Enhance the Urban Vitality Based on Graph Theory and Gravity Models (Case of District 9, Tehran) |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |