CN111553006A

CN111553006A - Large-scale rural residential layout generation design method in northern plain area

Info

Publication number: CN111553006A
Application number: CN202010327152.8A
Authority: CN
Inventors: 王江; 杨阳; 范伟; 赵伯伦; 赵继龙
Original assignee: Shandong Jianzhu University
Current assignee: Shandong Jianzhu University
Priority date: 2020-04-23
Filing date: 2020-04-23
Publication date: 2020-08-18
Anticipated expiration: 2040-04-23
Also published as: CN111553006B

Abstract

The invention discloses a large-scale rural residence layout generation design method in northern plain areas, and belongs to the field of digital urban and rural planning. The invention determines three-level generation rules of rural homesteads in northern plain areas: the level 1 rule describes a dwelling unit mode of 4-family combination, the level 2 rule describes a public space organization, a road network structure and a neighborhood mode of each 16-family combination, and the level 3 rule describes a land building density, a land internal and external relation and a community mode of each 64-family combination; building a homestead combination prototype of each level of rules through a Python programming language, automatically generating and screening samples generated by each level of rules, and building a corpus in npy format to store sample information; and for the dwg file of the given land parcel, identifying base information by means of a PyautoCAD module and automatically calling a sample from a corresponding sample pre-material library to generate a design. The invention realizes the automatic design of the layout of large-scale rural houses in northern plain areas.

Description

Large-scale rural residential layout generation design method in northern plain area

Technical Field

The invention relates to the field of digital urban and rural planning, in particular to a large-scale rural residential layout generation design method in northern plain areas.

Background

Township rural dwellings in china are usually designed and built on a large scale rapidly using standardized, "blueprint" design models. Excessive monotony, high vacancy rate, frequent over-reconstruction and the like. Now, the problem begins to spread to rural residences for 6 hundred million Chinese people, and the existing rural buildings are low in quality and have obvious urbanization and homogenization trends under the action of a conservative design method and a development mode.

From 2000, by effective planning and regulation, the rural settlement is reasonably planned on the premise of preferentially promoting urbanization, and the population of the rural area is promoted to appropriately and intensively live; on the basis, rural infrastructure needs to be reasonably configured, the quality of the living environment of the rural people is improved, and a space structure favorable for urban and rural integration is formed. When the strategy is implemented, local governments and development enterprises usually simulate the construction mode of urban residences, set up some top-down large-scale economically-applicable housing plans and projects, and uniformly develop large-scale housing units in a partition mode. In the process of design, construction and use, the rural houses are also regarded as 'residential machines' like urban houses, and through standardized and modularized 'copy' expression and efficient construction, the types and the styles of the rural houses can achieve high consistency and high degree of fitting with the environment, so that neat and uniform housing landscapes and relatively complete infrastructure can be quickly formed.

The shortcomings of the method are gradually exposed in the actual construction process:

1) increase in construction costs of houses and infrastructures;

2) the construction mode provided by the supplier is extensive and the industrialization degree is not high;

3) the traditional "blueprint" method adopted by architects is difficult to combine layout planning of large-scale residences and living needs of users in a limited design period, and further fails to provide design results with predefinition, stage, hierarchy and diversification.

Furthermore, traditional rural spaces in china, which originally feature dynamic, sustainable and spontaneous incremental construction, exhibit a clear tendency to urbanization and homogenization.

In order to solve the problems, a new generation of large-scale rural residential design mode in China needs to be explored, an efficient generation design method is explored for rural residences with small volume, large volume, diversity and complexity, and rural residential design is pushed to develop towards high-quality industrialization and differentiation.

Disclosure of Invention

In order to make up for the defects of the prior art, the invention provides a large-scale rural residential layout generation design method in northern plain areas. The coding rule is designed for large-scale rural residences in the plain areas in the northern China, so that the problem that the existing rural residences have obvious urbanization and homogenization trends in arrangement is solved.

The technical scheme of the invention is as follows:

a large-scale rural residence layout generation design method in northern plain areas comprises the following steps:

1) determining the basic modulus of the meshing of the homestead, and determining the dimension of the lane adopted in the design according to the average dimension of the lane; determining three-level rules and combined prototypes of all levels of rules;

2) encoding a primary rule by using a programming language, generating an initial corpus and a core corpus of a primary sample, and classifying the core corpus of the primary sample by using the programming language;

3) coding a secondary rule by using a programming language to generate an initial corpus, a core corpus and an extended corpus of secondary samples;

4) using a programming language to code a three-level rule, controlling a generation mode of a three-level sample and controlling and adjusting the density of a home base in the sample;

5) and automatically identifying a given land parcel through a computer, and calling and filling samples from the corpus generated under each level of rules.

As a preferable scheme, in the method for generating and designing the layout of the large-scale rural houses in the northern plain region, in the step 1),

determining the size of a standard homestead and the basic modulus of meshing the homestead at the design time according to the average size and the opening width of the homestead; determining the dimension of the lane adopted in design according to the average dimension of the lane;

determining a reduced type and three expanded type house bases by increasing or decreasing module units on the basis of the size of the standard house base;

using Python programming language, representing a homestead and a roadway by a number matrix; the lane in the matrix is represented by 0, and the home base is represented by 1;

the first-stage combined prototype comprises a dwelling unit-level sample consisting of 4 standard dwelling bases and a central cross road;

the secondary combination prototype comprises 4 primary samples, and neighbourhood combination primary samples are formed in a combination mode of Zone-A1/A2/A3/A4;

the three-level combined prototype comprises 4 second-level samples, a central cross lane and community-level samples formed in a Zone-B1/B2/B3/B4 combined mode.

As a preferable scheme, in the method for generating and designing the layout of the large-scale rural houses in the northern plain area, in the step 2),

calculating the number of combinations of units for respectively translating 4 homesteads by 1 lane in the established digital matrix of the primary sample combined prototype by using a Python programming language, calculating 79 forms in total, taking the 79 samples as a primary sample initial corpus, and storing the 79 samples in a primary sample initial corpus file by using an npy format file;

calling 79 samples stored in a primary sample initial corpus through a Python programming language, and accessing one by one through depth-first search to identify unreasonable samples and eliminate the unreasonable samples, wherein 16 samples are remained; using the 16 samples as a primary sample core corpus S0; storing the 16 samples in a primary sample core corpus file in an npy format file;

further, in step 2), the classification of the primary sample core corpus is specifically:

using Python programming language to call 16 samples stored in a primary sample core corpus file, identifying the condition that a road internally connected with 4 homesteads reaches four corner points of the primary sample, and dividing the 16 samples into four types: the lane can reach the Zone-A1 arranged at the northwest corner, the lane can reach the Zone-A2 arranged at the northeast corner, the lane can reach the Zone-A3 arranged at the southwest corner, and the lane can reach the Zone-A4 quadrant arranged at the southeast corner;

identifying whether a straight lane which is through up and down exists in the sample through a Python programming language, and dividing the samples in the Zone-A1, the Zone-A2, the Zone-A3 and the Zone-A4 into M and N subtypes, wherein the M type is a straight lane which is through up and down, and the N type is a straight lane which is not through up and down;

the classified samples are stored in the files of respective types in npy format files.

As a preferable scheme, in the step 3) of the large-scale rural residence layout generation design method in the northern plain area,

automatically calling samples from the files classified correspondingly according to four areas of Zone-A1/A2/A3/A4 in the digital matrix of the established secondary combined prototype for filling through a Python programming language;

using Python programming language to obtain mirror image combination of the samples according to the adjacent Zone; the M in the Zone-A1 and the M in the Zone-A3 cannot be arranged adjacently up and down; m types in the Zone-A2 and M types in the Zone-A4 cannot be adjacently arranged up and down for constraint, and 148 sample combination relations are calculated; storing the 148 samples in npy format in an initial corpus file of secondary samples;

and automatically calling the samples stored in the initial corpus file of the secondary sample by using a Python programming language for identification and screening, removing the samples of the lane which does not enter the central point of the secondary sample at the four boundaries in the secondary sample, and storing the remaining 102 samples in the core corpus file of the secondary sample in npy format.

Further, in the step 3), when constructing the extended corpus of the secondary sample, the extended corpus is divided into two types, wherein one type is the case that the standard homestead inside the sample is changed into the extended homestead; the other is the case of removing the home base from the sample.

As a preferable scheme, in the method for generating and designing the layout of the large-scale rural residence in the northern plain area, the standard homestead in the sample is changed into an extended corpus in the case of an extended homestead:

using Python programming language to call 102 samples of the secondary sample core corpus, adding the condition limit of the adjacent vehicle road, and then automatically controlling and expanding the size of a part of the homestead in the samples, wherein the expanded size of the homestead corresponds to the size of the 3 determined expanded homesteads; the method is divided into three categories according to different road approaching modes:

when a sample is adjacent to a vehicle road on one side, a random homestead which is not adjacent to the road and is positioned on a corner in the sample is used as a mark homestead, four homesteads adjacent to the road extend and expand 1 modulus to the vehicle road, two homesteads at the end points of opposite sides of the edge which is perpendicular to the vehicle road of the mark homestead extend and expand 1 modulus to the outside simultaneously, and 408 sample forms are formed after calculation by using a Python programming language; storing the 408 samples in a T1 file in npy format;

when the two sides of the sample face the vehicle and travel, homesteads which are not in the sample and are positioned at the corners are selected as the marked homesteads, 7 adjacent homesteads extend to the vehicle and travel by 1 modulus, and the two homesteads are calculated by using a Python programming language and then form 408 sample forms; storing the 408 samples in a T2 file in npy format;

when the sample does not face the vehicle, randomly selecting a homestead at a corner as a marked homestead, expanding 1 modulus to one side from the homestead at 2 corner points adjacent to the marked homestead, expanding 1 modulus to two sides from the diagonal homestead of the marked homestead, and forming 408 sample forms after calculation by using a Python programming language; the 408 samples were stored in the T3 class file in npy format.

As a preferable scheme, the northern plain area large-scale rural residential layout generation design method removes an extended corpus under the condition of a sample home base:

when a sample approaches a road, calling a T1-type sample by using a Python programming language, automatically identifying and marking 8 homesteads and lanes in two zones with the homesteads perpendicular to a driving road, reserving, removing the remaining 8 homesteads and lanes, calculating to generate 408 samples, and storing the 408 samples in S1-type files in a npy format;

when the sample is not adjacent to a road, calling a T3 sample by using a Python programming language, identifying and reserving 8 homesteads and lanes in two zones on the transverse edge or the longitudinal edge of the labeled homestead in the sample, removing the remaining 8 homesteads and lanes, and generating 816 samples after calculation; storing the 816 samples in an npy format in an S2 file;

calling a T3 sample by using a Python programming language, identifying 4 butts and lanes marked on the transverse edge or the longitudinal edge of the butts in the sample, reserving the samples, removing the remaining 12 butts and lanes, and calculating 36 samples after the same sample is removed; the 36 samples were stored in the S3 class file in npy format.

As a preferable scheme, in the method for generating and designing the layout of the large-scale rural houses in the northern plain area, in the step 4), the three-level samples are generated in a manner,

and calling a secondary sample from the secondary sample extended corpus for filling according to the tertiary sample prototype by using a Python programming language to generate a tertiary sample.

Further, in the step 4), the controlling and adjusting of the density specifically includes:

calling samples in the secondary sample core corpus and the secondary sample extended corpus by using a Python programming language, controlling the condition of deleting the homesteads in the called samples at different densities by using the Python programming language, and deleting the homesteads adjacent to the driving roads to open an entrance and an exit of a community; and deleting the homesteads at other positions to enlarge the open space in the community.

and automatically identifying the standard homestead which is not deleted and becomes the public space in the four homesteads in the sample center subjected to the subtraction processing by using a Python programming language, reducing the standard homestead by 1 module, wherein the size of the reduced homestead corresponds to the size of the determined reduced homestead.

As a preferable scheme, the method for generating and designing the layout of the large-scale rural houses in the northern plain area comprises the following steps of 5):

for a given dwg-format base file, base boundaries, vehicle roads and internal exclusion layers in Dwg format are identified by means of a Python module, then samples in a corresponding attribute sample library are randomly called by using a Python programming language, and the home base is deleted according to the sample density controlled by coding, the size of the home base is reduced, and then a CAD (computer-aided design) image division layer is automatically controlled to fill the samples in the base.

The invention provides a large-scale residential generation design method based on a shape grammar, which is used for meeting the construction requirements of rural residences which are massively emerged in plain areas in the northern China and solving the problems of obvious tendency of urbanization and homogenization of the existing rural residences due to the conservation of design methods and development modes. By investigating the traditional villages in northern plain areas and discussing the overall layout, form and characteristics of village houses, the three-level generation rule of the house-base combination is determined: the level 1 rule describes a dwelling unit pattern of 4-user groups, the level 2 rule describes a public space organization, a road network structure and a neighborhood pattern of every 16-user group, and the level 3 rule describes a land building density, a land internal and external relation and a community pattern of every 64-user group. Building a homestead combination prototype of each level of rules through a Python programming language, automatically generating and screening samples generated by each level of rules, and building a file library in npy format to store sample information; for a dwg file for a given parcel, base information may be identified with the aid of a PyautoCAD module and samples automatically called from a corresponding sample library to generate a design.

The invention has the beneficial effects that:

the invention is an effective method for generating large-scale rural residential planning in northern plain areas; and using each level of sample library of Python coding, identifying the base form of the dwg file through Python, automatically calling samples with corresponding attributes by Python for filling, and controlling the building density of the samples during filling through a coding language in the filling process.

The invention realizes automatic design, improves design efficiency, can generate a plurality of different schemes for comparison in a short time by randomly calling samples in computer languages, and realizes diversity of scheme generation, thereby helping designers to rapidly generate schemes. Moreover, the diversification of the combination forms of the home bases in all levels of samples also enables the traditional rural texture to be continued, and avoids the urbanization problem of large-scale rural residential layout design.

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 only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.

FIG. 1 is a standard homebase, a reduced homebase, an enlarged homebase profile, and a building envelope;

FIG. 2 is a three level standard type of homestead portfolio;

FIG. 3 is an example of 79 samples in a primary sample initial corpus;

FIG. 4 is a graph of the remaining 16 satisfactory sample types after 79 samples have been screened for unsatisfactory samples;

FIG. 5 is a chart of classifying sample types satisfying the requirements in 16;

FIG. 6 is an example of 148 secondary samples formed after primary samples are combined according to a secondary sample prototype and an example of unsatisfactory samples;

FIG. 7 is an example of 148 samples further screened for unsatisfactory samples and satisfactory samples;

FIG. 8 is a sample example of class T1;

FIG. 9 is a sample example of class T2;

FIG. 10 is a sample example of class T3;

FIG. 11 is a diagram illustrating the generation of a S1 sample;

FIG. 12 is a diagram illustrating the generation of a S2 sample;

FIG. 13 is a sample example of class S3;

FIG. 14 is an example of a three-level sample morphology after a two-level sample is combined according to a three-level prototype;

FIG. 15 is an example of a three level sample morphology at medium density requirements;

FIG. 16 is an example of a 1-modulus reduction of a standard homeland that does not have 4 homelands deleted from each secondary sample center to a common space under medium density requirements;

fig. 17 is a given base plot;

FIG. 18 shows the result of the identification and padding in example 1.

Detailed Description

firstly, determining a basic modulus of meshing of a home base, and determining the dimension of a lane adopted in design according to the average dimension of the lane; determining three-level rules and combined prototypes of all levels of rules;

secondly, a primary rule is coded by using a Python programming language, an initial corpus and a core corpus of the primary sample are generated, and the core corpus of the primary sample is classified by using the Python programming language;

then, a Python programming language is used for coding a secondary rule to generate an initial corpus, a core corpus and an extended corpus of secondary samples;

then, a third-level rule of a Python programming language code is used for controlling the generation mode of the third-level sample and controlling and adjusting the density of the home base in the sample;

and finally, automatically identifying the given land parcel through a computer, and calling samples from the corpus generated under each level of rules for filling.

Specifically, the method comprises the following steps:

for the homebase size and road size, the standard homebase size and basic modulus of the homebase meshing at design time are determined from the average size of the homebase and the opening width of the village under study. On the basis of the standard size of the house, in order to realize diversified selection of the house size, 1 kind of reduced house and 3 kinds of enlarged house are determined by increasing or reducing module units, and the size of the lane used in design is determined according to the average size of the lane in the village. In order to realize the coding representation of the home and the lane, the home and the lane are represented in a digital matrix form by using a Python programming language, the lane in the matrix is represented by 0, and the home is represented by 1.

For a 3-level combination prototype of the home, in order to meet the requirement of a hierarchy level in a newly-built residential area, a standard home combination type digital matrix of three hierarchy levels is established by using a Python programming language. The first-stage combined prototype comprises a dwelling unit-level sample consisting of 4 standard dwelling bases and a central cross road; the two-stage combined prototype comprises four first-stage samples, 16 standard homesteads and neighborhood combined level samples formed in a combination mode of Zone-A1/A2/A3/A4; the three-level combined prototype comprises four second-level samples, 64 standard homesteads, a central cross road and community-level samples formed in a Zone-B1/B2/B3/B4 combined mode.

An initial corpus and a core corpus for the primary samples. In order to provide diversified and alternative samples, the Python programming language is used to calculate the number of combinations of units which respectively translate four bases by 1 lane in the established digital matrix of the primary sample combination prototype, and then 79 types of forms are calculated in total, and the 79 types of samples are used as a primary sample initial corpus. To automatically retrieve the samples of the initial corpus, the computer stores the 79 samples in the initial corpus file of the primary sample in npy format. In order to meet the requirement that a road is arranged in the primary sample and can directly pass through four homesteads, and the courtyards of the four homesteads are not arranged on the north side of the main house and the mountain wall surface of the main house, 79 samples stored in the primary sample initial corpus are called through a Python programming language to be accessed, unreasonable samples are identified and eliminated, and 16 samples are remained. The 16 samples are used as a primary sample core corpus S0, and the 16 samples are stored in a primary sample core corpus file in a npy format file in order to realize that a computer can automatically call the samples in the core corpus.

For the classification of a primary sample corpus, in order to realize that each primary sample home can reach the central area of a secondary sample when the primary samples are combined into the secondary samples, the Python programming language is used for calling 16 samples stored in a primary sample corpus file, the condition that the lane of four internally connected home bases reaches four corners of the primary sample is identified, and the 16 samples are divided into four types: the lane can reach the Zone-A1 arranged at the northwest corner, the lane can reach the Zone-A2 arranged at the northeast corner, the lane can reach the Zone-A3 arranged at the southwest corner, and the lane can reach the Zone-A4 quadrant arranged at the southeast corner, because the case that the lane can reach two corner points of the rectangular outline exists in the sample, 5 samples are respectively arranged in each type after the summary. In addition, in order to avoid the problems of monotonous interface caused by too monotonous and long straight lane which may be generated during the combination of the primary samples in advance, and the like, whether a straight lane which penetrates up and down exists in the samples is identified through a Python programming language, samples in the Zone-A1, the Zone-A2, the Zone-A3 and the Zone-A4 are divided into two sub-types of M and N, wherein the M is a straight lane which penetrates up and down, and the N is a straight lane which does not penetrate up and down. In order to realize that the computer can automatically call samples of different classifications, the classified samples are stored in files of respective types in npy format.

For the initial corpus of the secondary samples, the Python programming language is used, and the samples are automatically called from the files of the corresponding classification according to four areas of Zone-A1/A2/A3/A4 in the established numerical matrix of the secondary combined prototype to be filled. In order to avoid the problem of monotonous space caused by combination in advance, mirror combination of samples cannot occur according to adjacent zones by using a Python programming language; the M in the Zone-A1 and the M in the Zone-A3 cannot be arranged adjacently up and down; m classes in Zone-A2 and M classes in Zone-A4 cannot be adjacently arranged up and down for constraint, 148 sample combination relations are calculated, and in order to realize that a computer can automatically call the 148 samples, the 148 samples are stored in an initial corpus file of a secondary sample in a npy format.

For the secondary sample core corpus, in order to solve the problem that no roadway can enter the central point of the secondary sample exists at four boundaries of the secondary sample, a Python programming language is used for automatically calling the sample stored in the secondary sample initial corpus file for identification and screening, and only 102 samples can meet the requirement after the sample with the condition is removed. In order to realize that the computer can automatically call 102 samples meeting the requirements, the 102 samples are stored in the core corpus file of the secondary sample in npy format.

In order to adapt to the geometric boundary of a residential plot and enrich the types of the homesteads, a secondary sample expansion corpus is constructed and divided into two types: 1. the case where the intra-sample standard homebase becomes the enlarged homebase. 2. The case of the home base in the sample is removed.

Under the condition that the standard homestead in the sample is changed into the expanded homestead, 102 samples of the secondary sample kernel corpus are called by using a Python programming language, the size of the partial homestead in the sample is automatically controlled to be expanded after the condition limitation of the adjacent vehicle road is added, and the expanded homestead size corresponds to the 3 determined expanded homestead sizes. The method is divided into three categories according to different road approaching modes:

1. when a sample is adjacent to a vehicle road on one side, a random homestead which is not adjacent to the road and is positioned on a corner in the sample is used as a mark homestead, four homesteads adjacent to the road extend and expand 1 modulus to the vehicle road, two homesteads at the end points of opposite sides of the edge which is perpendicular to the vehicle road of the mark homestead extend and expand 1 modulus to the outside simultaneously, and 408 sample forms are formed after calculation by using a Python programming language; the 408 samples were stored in the T1 class file in npy format.

2. When the sample is on the adjacent vehicle road at two sides, in order to control the contrast relation when the neighborhoods are expanded, the neighborhoods which are not adjacent to the road and are positioned at the corners in the sample are selected as the mark neighborhoods, 7 neighborhoods adjacent to the road extend and expand 1 modulus to the vehicle road, and the 408 sample forms are formed after calculation by using Python programming language. To facilitate the automatic retrieval of the 408 samples by the computer, the 408 samples are stored in the T2 class file in npy format.

3. When the sample is not adjacent to the vehicle road, in order to control the contrast relation when the homestead is expanded, a homestead at a corner is randomly selected as a marked homestead, the homestead at 2 corners adjacent to the marked homestead is expanded to a single side by 1 modulus, the opposite corners of the marked homestead are respectively expanded to two sides by 1 modulus, and the sample forms 408 sample forms after calculation by using Python programming language. To facilitate the automatic retrieval of the 408 samples by the computer, the 408 samples are stored in the T3 class file in npy format.

The case of removing the home base in the sample is classified into 3 types:

And a generation mode of the three-level sample is that the Python programming language is used for calling the two-level sample from the two-level sample extended corpus to be filled according to the three-level sample prototype, so that the three-level sample is generated. And a new sample library is not required to be established because the number of the three-level samples is excessive and the three-level samples are not required to be combined continuously to generate higher-level samples. Therefore, the third-level samples can be realized by randomly calling the second-level sample combination by using Python language when the plot is filled.

In order to enable the three-level samples to have a certain level of public open space (or green space) and reduce the building density, density control adjustment is required, samples in a secondary sample core corpus and a secondary sample extended corpus are called by using a Python programming language, the condition that the homesteads are deleted in the called samples at different densities is controlled by using the Python programming language, and the homesteads adjacent to a vehicle road are deleted and are used for opening an entrance and exit of a community; and deleting the homesteads at other positions to enlarge the open space in the community.

In order to further increase the open space of the center position of the secondary sample, a Python programming language is used for automatically identifying the standard homestead which is not deleted and becomes the public space in the four homesteads of the sample center subjected to subtraction processing, the standard homestead is reduced by 1 module, and the size of the reduced homestead corresponds to the size of the determined reduced homestead. The density control process, when filling a given plot, uses Python programming language to automatically reduce the number of and size of the homesteads when calling samples, so that a new sample library does not need to be established.

For a given dwg-format base file, base boundaries, a vehicle road and an internal exclusion layer in Dwg format are identified by means of a module (module) PyautoCAD of Python, samples in a corresponding attribute sample library are randomly called by using Python programming language, and then the samples are deleted according to the sample density controlled by coding, the size of the home is reduced, and then the CAD sub-map layer is automatically controlled to fill the sample in the base. For example, when a land block is adjacent to a traffic area, samples in T1, T2 and S1 are automatically called to adjust the density and then are filled; if the land is not adjacent to the traffic road area, automatically calling samples in T3 and S2, adjusting the density and filling; the remaining smaller regions are filled by calling the samples in S0, S3.

Example 1

Taking wash village of Laiyang city, Shandong province as an example, the size of the home and the road size are determined by conducting investigation and drawing investigation conclusion. The form type of the country homestead in the Shandong plain area is more square, the major aspects are rectangle and square, and the room opening size is between 3 and 3.3 m; influenced by climate and civil convention, the main house faces south in north, the number of layers does not exceed 2, and the entrance and exit are not suitable to be arranged on the gable wall and the north boundary of the courtyard.

The road is divided into two levels of a vehicle road and a roadway road, wherein the former is used for the passing of household motor vehicles and agricultural vehicles and has the width of about 6m, and the latter is used for walking and has the width of between 3 and 4 m; the lane is mainly a folded line and an end line, and the number of straight through paths is less; the area of the home base adjacent to the vehicle is large because it is used for business services.

Based on the above, 3.3 meters is selected as the basic module of the homestead meshing, 6.6 meters is selected as the width of the roadway, and 3.3 meters is selected as the width of the roadway, and the Python programming language is used for coding.

The standard homestead has an aspect ratio of 13.2m to 9.9m and an area of 130.68 square meters. Considering the adaptability of the homesteads in the residential plots, 1 kind of reduced type and 3 kinds of enlarged type homesteads are additionally determined, and refer to fig. 1. Regardless of the type of home base, the top two rows of squares are occupied by a floor plan of the main building, and the maximum outline range is shown by the diagonal lines in fig. 1.

Referring to fig. 2, three levels of standard type home base combination types are determined:

the first level is composed of 4 homesteads and a central cross lane to form an initial sample;

the second level comprises 4 first-level samples and 16 homesteads, and initial samples of neighborhood relations are formed in a combination mode of Zone-A1/A2/A3/A4;

the three stages comprise 4 secondary samples, 64 homesteads and a central cross road, and the initial samples of the residential area are formed in a combination mode of Zone-B1/B2/B3/B4.

And sequentially establishing a number matrix representation by using the 3-level combination type in a Python programming language.

Referring to fig. 3, using Python programming language to calculate, respectively translating four homesteads by 1 modulus unit in the established digital matrix of the primary sample combination prototype, a total of 79 combination relations occur, and the 79 samples are stored in the primary sample initial corpus file in npy format files.

In order to ensure that one lane can directly pass through four homesteads in the first-stage sample, and the courtyards of the four homesteads are not arranged on the north side of the main house and the mountain wall surface of the main house, the sample of the first-stage sample initial corpus is taken through Python and identified, and samples which are not satisfied in the initial corpus are excluded, for example, the homestead at the upper left corner of the sample A in the figure 4 does not satisfy the requirement because no internal lane is communicated, and the courtyard of the homestead at the lower right corner of the sample B can only be arranged on the mountain wall and does. After screening, only 16 samples can satisfy the requirement, and the 16 samples and the primary sample kernel corpus S0 are stored in an npy format file, as shown in fig. 4.

Referring to fig. 5, in order to realize that each of the first-level samples can reach the center area of the second-level sample when the first-level samples are combined into the second-level sample, 16 samples stored in the first-level sample core corpus file are called by using a Python programming language, the condition that the roadway internally connecting the four bases reaches the four corners of the first-level sample is identified, and the 16 samples are classified into four types: the lane can reach the northwest corner and is placed in the Zone-A1, the lane can reach the northeast corner and is placed in the Zone-A2, the lane can reach the southwest corner and is placed in the Zone-A3, and the lane can reach the southeast corner and is placed in the Zone-A4, and 5 samples are respectively arranged in each type after induction.

In addition, in order to avoid the problems of monotonous interface caused by too monotonous and long straight lane which may be generated in the first-level sample combination, the Python language is used for identifying whether the straight lane which penetrates up and down exists in the sample, and the straight lane is divided into two subtypes of M and N, wherein M is a straight lane which penetrates up and down, and N is a straight lane which does not penetrate up and down.

And (3) automatically calling samples from the files of the corresponding classification according to four areas of Zone-A1/A2/A3/A4 in the digital matrix of the established two-level combined prototype for filling by using a Python programming language. In order to avoid the problem of monotonous space caused by combination in advance, mirror combination of samples cannot occur according to adjacent zones by using a Python programming language; the M in the Zone-A1 and the M in the Zone-A3 cannot be arranged adjacently up and down; the M class in the Zone-a2 and the M class in the Zone-a4 cannot be adjacently arranged up and down for constraint, a total of 148 sample combination relations are calculated, and the 148 samples are stored in an initial corpus file of a secondary sample in a npy format, see the example in fig. 6.

Referring to fig. 7, in order to solve the problem that there are no lanes that can enter the central point of the secondary sample at the four boundaries of the secondary sample, 148 samples of the initial corpus of the secondary sample are automatically called using Python language, and samples that have no lane that can enter the central point of the secondary sample at one side are identified and excluded, for example, two lanes above the sample a shown in fig. 7 cannot enter the central point and thus do not meet the requirement. It was calculated that only 102 samples could satisfy the requirement, and the 102 samples were stored in the core corpus file of the secondary sample in npy format.

In order to adapt to the geometric boundary of the residential plot and enrich the types of the homesteads, an extended corpus of secondary samples is constructed. The extended corpus is divided into two categories: 1. the case where the intra-sample standard homebase becomes the enlarged homebase. 2. The case of the home base in the sample is removed.

For the case that the standard house inside the sample is changed into the enlarged house, the condition that 102 samples of the secondary sample core corpus are added into the temporary driving path is limited, and the condition is divided into three types:

1. when a sample is adjacent to a vehicle road on one side, a random homestead which is not adjacent to the road and is positioned on a corner in the sample is used as a mark homestead, four homesteads adjacent to the road extend and expand 1 modulus to the vehicle road, two homesteads at the end points of opposite sides of the edge which is perpendicular to the vehicle road of the mark homestead extend and expand 1 modulus to the outside simultaneously, and 408 sample forms are formed after calculation by using a Python programming language; defining this as the T1 class, the 408 samples are stored in the T1 class file in npy format, as shown in fig. 8.

2. When the sample is traveling along both sides of the vehicle, the home at the corner, which is not adjacent to the road in the sample, is selected as the labeled home, and 7 homes adjacent to the road are extended by 1 module toward the vehicle, which is defined as T2, as shown in fig. 9.

3. When the sample is not adjacent to the traffic, one of the homes at the corner is randomly selected as the marked home, 2 homes adjacent to the corner are expanded by 1 module to one side, and the diagonal home is expanded by 1 module to both sides, and this is designated as T3, as shown in fig. 10.

Through the process of Python programming language coding, 102 samples in the secondary sample core corpus are automatically called, corresponding expansion sample forms are calculated and generated, and then expansion samples are respectively stored in corresponding files by using npy format files.

For the case of removing the home base from the sample, 3 categories are classified:

1. when the sample approaches the road, 8 homesteads and lanes in two zones perpendicular to the driving road of the marked homestead in the T1 sample are reserved, the remaining 8 homesteads and lanes are removed, and the type is determined as S1. Referring to fig. 11, the dotted line is the reserved homestead and the removal of the homestead outside the lane dotted line.

2. When the sample is not adjacent to the road, the 8 neighborhoods and the lanes on the transverse side or the longitudinal side of the labeled homeland in the T3 sample are reserved, the remaining 8 homeland and the lanes are removed, and the type is determined as S2. Referring to fig. 12, the dotted line is the remaining home and lane removed outside the dotted line.

3. The 4 butts and lanes on the lateral or longitudinal side of the labeled butts in the T3-type sample are retained, the remaining 12 butts and lanes are removed, and this is designated as S3, as shown in fig. 13. And coding the process by a Python programming language, automatically calling the samples in the corresponding expansion library, calculating and generating the corresponding expansion sample form, and respectively storing the newly generated expansion samples in the corresponding files by using npy format files.

Referring to fig. 14, an ideal form containing 64 homesteads is formed by combining 4 secondary samples in a three-stage prototype mode, subject to the size of the plot and the randomness of filling. Firstly, whether the adjacent main road is existed or not and the number of the adjacent main road sides are automatically identified through a Python programming language, and after identification, the expansion sample is randomly called from the corresponding T1, T2 and T3 libraries.

In order to enable the tertiary sample to have a certain level of public open space (or green space), subtraction processing needs to be carried out on the homesteads in the secondary sample, and if the homesteads adjacent to the main road are deleted, the subtraction processing is used for opening the entrance and exit of the community; if the homesteads at other positions are deleted, the method is used for expanding the open space in the community. The corresponding subtraction method is defined below according to the density. Table 1 shows the corresponding homestead removal constraints under different density requirements; fig. 15 is an example of three-level samples in the medium-density format, where the homestead filled in the diagonal grid is the removed homestead, the density control shown in table 1 is coded into Python language, and the Python code during filling with different densities is adjusted to control the sample density.

TABLE 1 Hotel removal constraints for different density requirements

And then, the Python programming language is used for automatically reducing the standard homesteads which are not deleted and become the public spaces in the 4 homesteads in the secondary sample center by 1 module in the length direction, so that the open space of the secondary sample center is further increased, and the homesteads shown by the broken line frame shown in fig. 16 are in the form of the reduced homesteads.

Referring to fig. 17, for a given block, the module (module) PyautoCAD of Python is used to identify the base boundary, the vehicle road and the exclusion layer of the factory area in dwg format, and then the PyautoCAD calls the samples in the corresponding category sample library to identify and fill the samples according to the block attributes (base boundary and vehicle road), respectively, as shown in fig. 18. And during filling, the filling probability of each level sample is set through the Python programming language to control the generated result.

Finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; such modifications and substitutions do not depart from the spirit and scope of the present invention, and they should be construed as being included in the following claims and description.

Claims

1. A large-scale rural residence layout generation design method in northern plain areas is characterized by comprising the following steps:

2. The northern plains large-scale rural residential layout generation design method of claim 1, wherein: in the step 1), the step (A) is carried out,

3. The northern plains large-scale rural residential layout generation design method according to claim 1 or 2, characterized in that: in the step 2), the step (c) is carried out,

calculating the number of combinations of units for respectively translating 4 homesteads by 1 lane in the established digital matrix of the primary sample combined prototype by using a Python programming language, calculating 79 layout forms in total, taking the 79 samples as a primary sample initial corpus, and storing the 79 samples in a primary sample initial corpus file by using an npy format file;

calling 79 samples stored in a primary sample initial corpus through a Python programming language, and accessing one by one through depth-first search to identify unreasonable samples and eliminate the unreasonable samples, wherein 16 samples are remained; using the 16 samples as a primary sample core corpus S0; the 16 samples are stored in the primary sample core corpus file in npy format.

4. The northern plain area large-scale rural residential layout generation design method according to claim 3, wherein in step 2), the classification of the primary sample core corpus is specifically:

5. The northern plains large-scale rural residential layout generation design method according to claim 1 or 2, characterized in that: in the step 3), the step (c),

and automatically calling the samples stored in the initial corpus file of the secondary sample by using a Python programming language for identification and screening, removing lane samples of which four boundaries cannot enter the central point of the secondary sample from the secondary sample, and storing the remaining 102 samples in the core corpus file of the secondary sample in npy format.

6. The northern plains large-scale rural residential layout generation design method of claim 5, wherein: in the step 3), when an extended corpus of the secondary samples is constructed, the extended corpus is divided into two types, wherein one type is the condition that the standard homestead inside the samples is changed into an expanded homestead; the other is the case of removing the home base from the sample.

7. The northern plains large-scale rural residential layout generation design method according to claim 6, wherein the standard homestead inside the sample is changed to an extended corpus in case of an extended homestead:

8. The northern plains large-scale rural residential layout generation design method of claim 7, wherein the extended corpus in the case of the sample home base is removed:

9. The method for generating and designing the layout of large-scale rural houses in northern plains according to any one of claims 6 to 8, wherein in the step 4), the three-level samples are generated in a manner,

10. The northern plain area large-scale rural residential layout generation design method according to claim 9, wherein in step 4), the density control adjustment is specifically: