A terminal-oriented house type customization automatic derivation method and system
By establishing an apartment layout matching recommendation algorithm model and a virtual trial stay experience, the problem of apartment layout customization being greatly influenced by the designer's subjectivity has been solved, achieving efficient and accurate personalized apartment layout customization and improving customization efficiency and matching accuracy.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHINA CONSTR THIRD ENG BUREAU GRP CO LTD
- Filing Date
- 2026-03-05
- Publication Date
- 2026-06-12
AI Technical Summary
Existing apartment layout customization designs are heavily influenced by the designer's subjectivity, making it difficult to accurately identify customer needs. The process is complex and inefficient, resulting in data that is reasonable but provides a poor user experience, or a good user experience that does not meet core requirements.
The system adopts an automated derivation method for customized apartment layouts, which involves establishing an apartment layout adaptation and recommendation algorithm model, collecting users' original verbal data in layers, generating virtual trial apartment layout templates, organizing trial stay experience annotation and feedback tags, adjusting the apartment layout template adaptation specifications, forming a two-dimensional customization basis, and finally iteratively modifying the customized apartment layout plan until it meets the user's needs.
It achieves precise positioning of user needs, eliminates subjective judgment of designers, makes large-scale customization more efficient, has high customization efficiency, unified recommendation standards, more comprehensive personalized adaptation, integrates trial stay experience and original needs, and makes the plan more balanced and more in line with actual living.
Smart Images
Figure CN122199099A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of apartment layout customization technology, specifically to an automated derivation method and system for apartment layout customization for end users. Background Technology
[0002] In today's era, everyone pursues individuality and has increasingly higher requirements for apartment design. In the process of customization, there are more personalized, artistic, and functional needs. At the same time, with the development of Internet technology and artificial intelligence technology, the impact of the Internet on the apartment design industry in terms of convenience, openness, low cost, and high efficiency is becoming increasingly prominent.
[0003] Current apartment layout customization involves clients describing their needs, followed by designers creating the design. This method is heavily influenced by the designer's subjectivity, relying solely on experience and failing to accurately pinpoint client requirements, making the process complex. Traditional apartment layout customization relies on manual matching by designers, which is not only inefficient but also leads to different designers recommending different layouts for the same needs due to varying levels of experience. Furthermore, traditional apartment layout customization either focuses solely on data or solely on user experience, resulting in data that is reasonable but offers a poor user experience, or a good user experience that does not meet core needs. Summary of the Invention
[0004] This invention proposes an automated derivation method and system for terminal-oriented apartment layout customization to solve the technical problems mentioned in the background.
[0005] To achieve the above objectives, the present invention adopts the following technical solution: An automated derivation method for customized apartment layouts for end-users, comprising the following steps: S1: Establish a housing type adaptation recommendation algorithm model; S2: Collect users' raw spoken language data in layers and convert the raw spoken language data into housing demand data, which includes basic housing demand, functional preference demand and personalized customization demand; S3: Substitute the residential demand data into the apartment type adaptation recommendation algorithm model to generate a virtual trial apartment template and customized apartment type adaptation specifications for the user. S4: Organize users to experience the virtual trial living room model through a three-dimensional virtual space, and mark the trial living feedback label of the model according to the experience. The trial living feedback label includes the adaptation item, the optimization item, and the rejection item. S5: Adjust the adaptation specifications of the apartment template based on the trial stay feedback tags, and combine them with the customized apartment adaptation specifications to form a two-dimensional customization basis, and customize the finished apartment plan accordingly. S6: The user experiences the customized finished apartment plan as a physical sample, judges its suitability for living, and obtains the experience results; S7: Iteratively modify the customized apartment layout specifications based on the experience results until the user's needs are met.
[0006] Preferably, S1 includes: S11: Classify and enter the basic attributes of the apartment type and the corresponding layout attributes into the database for archiving, collect complete living needs details of a large number of sample households, and form a sample living needs dataset; S12: Organize sample residents to experience various apartment layout options with different sizes and record their feedback on the size and layout of the apartments. S13: Calculate the redundancy parameters of the apartment layout based on sample residential demand data, suitable apartment size and apartment layout type; S14: Use the sample housing demand data as experience samples, divide them into training set and test set according to a preset ratio, and build a deep learning neural network algorithm model. S15: Input the sample housing demand data, corresponding suitable apartment size and apartment layout type into the model for training, continuously optimize the model parameters until the preset adaptation accuracy requirements are met, and finally construct an apartment adaptation recommendation algorithm model that includes a size recommendation model and a layout type recommendation model. The formula for calculating the redundancy parameter of the apartment layout is as follows:
[0007] in, For the first The first type of apartment layout Redundancy parameters in apartment layout corresponding to various apartment sizes. For the first The first type of apartment layout Average standard living requirements corresponding to various apartment sizes For the first The first type of apartment layout Actual apartment layout data corresponding to various apartment sizes. The number of sample households For the first The first type of apartment layout The first type of apartment size corresponds to the Housing needs data for a sample of households.
[0008] Preferably, S13 includes: S131: Classify the sample residential demand data according to the type of apartment layout that is suitable for the apartment type, and further subdivide the same layout type according to the size of the apartment type. S132: Calculate the mean of sample living demand data corresponding to each type of apartment size, and perform linear fitting on the mean to obtain the fitted data; S133: Based on the fitted data, derive the standard living requirements data corresponding to different apartment sizes under each apartment layout type; extract the actual apartment layout data corresponding to each apartment layout type and each apartment size; S134: Calculate the sample dispersion of the difference between standard residential demand data and actual unit layout data, and calculate the redundancy parameter of unit layout. The formula for calculating the mean of the sample housing demand data is as follows:
[0009] in, For the first The demand weights of each sample household.
[0010] Preferably, adjusting the adaptation specifications of the apartment layout model based on trial stay feedback tags includes the following steps: S51: Determine whether the space lighting requirements in the user's residential needs data are within the preset lighting requirements threshold range, and adjust the size and layout type of the apartment model according to the lighting adjustment coefficient; S52: Determine whether the functional area width requirement in the user's residential needs data is within the preset functional area width threshold range, and adjust the size and layout type of the apartment template accordingly through the functional area width adjustment coefficient; S53: Determine whether the user's residential needs data on circulation smoothness is within the preset circulation smoothness threshold range, and adjust the size and layout type of the apartment model according to the circulation optimization coefficient; S54: Determine whether the living area requirement in the user's living demand data is within the preset area requirement threshold range, and adjust the size and layout type of the apartment template accordingly through the area adaptation coefficient; S55: Based on the results of four adjustments, determine the size and layout type of the apartment model that precisely matches the user's living needs.
[0011] Preferably, the process of creating a customized finished apartment plan based on the adjusted apartment template specifications and the customized apartment specifications obtained from user living needs data includes the following steps: Based on the preset mapping rules between customized apartment types and apartment templates, the adjusted apartment template dimensions and layout types are converted into the corresponding customized apartment basic dimensions and basic layout types. The target size and layout type of the customized apartment are directly derived from the user's living needs data; the base size and target size of the customized apartment are weighted and summed using the final apartment parameter weighted summation formula to obtain the final apartment size; Similarly, the weighted summation formula for the final apartment type parameters is used to weight and fuse the basic layout type of the customized apartment type with the target layout type to determine the final apartment type. Based on the final apartment dimensions and layout type, complete the design and production of the customized finished apartment plan; The weighted summation formula for the final apartment layout parameters is as follows:
[0012] in, For the final apartment layout parameters, Based on the weights of the basic parameters, To customize the basic parameters of the apartment layout, To customize the target parameters for the apartment type.
[0013] Preferably, the process of establishing a unit type adaptation recommendation algorithm model, which includes a size recommendation model and a layout type recommendation model, includes the following steps: First, build a size recommendation model independently, then build a layout type recommendation model separately. After both models meet the preset adaptation accuracy requirements, integrate and optimize them to form a complete apartment type adaptation recommendation algorithm model.
[0014] Preferably, the establishment of the size recommendation model includes the following steps: Extract the redundant parameters of the apartment layout corresponding to the sample housing demand data, calculate the difference between the sample housing demand data and the corresponding redundant parameters, and obtain the corrected new housing demand data. Interpolation is used to match the area requirements in the new housing demand data with the basic area in the apartment layout data to obtain the preliminary apartment dimensions corresponding to the area requirements. By using matrix similarity analysis, the spatial dimension data in the new housing demand data is compared with the functional area dimension data in the apartment layout data to obtain the reference apartment dimensions corresponding to the spatial dimensions; The sample housing demand data and corresponding suitable apartment sizes in the training set are input into the established deep learning neural network algorithm model for training. The model parameters are continuously adjusted until the first preset adaptation accuracy is met, and the apartment sizes derived by the model are obtained. Using a linear regression algorithm, the preliminary apartment size, reference apartment size, model-derived apartment size, and sample-fitted apartment size are integrated. The deep learning neural network algorithm model is further optimized until the second preset fitting accuracy is met, and finally a size recommendation model is formed.
[0015] Preferably, the conversion of raw spoken data into residential demand data includes: Establish a mapping library between common customer language and professional apartment terminology. The mapping library contains common customer colloquial needs and their corresponding unique professional expressions. The collected raw spoken data from users is matched sentence by sentence with common expressions in the mapping library to initially generate corresponding professional requirement expressions; Professional designers were assigned to manually verify the initial converted professional requirements descriptions, and, based on user supplementary explanations, correct any deviations in the descriptions and supplement any missing details. The verified professional requirements are quantified into professional data that can be entered into the model, and then integrated with the basic requirements data filled in on paper to form professional residential requirements data.
[0016] An automated apartment layout customization system for end users, comprising: The model building module is used to build a model for apartment type matching recommendation algorithms. The data acquisition and processing module is used to collect users' raw spoken language data in layers and convert the raw spoken language data into residential demand data; The virtual template generation module is used to input the residential demand data into the apartment type adaptation recommendation algorithm model to generate a virtual trial apartment type template and customized apartment type adaptation specifications for the user. The tag generation module is used to organize users to experience virtual trial living room models in a 3D virtual space and to label the trial living room models with trial living feedback tags based on their experience. The apartment layout customization module is used to adjust the adaptation specifications of the apartment layout template based on the trial stay feedback tags. Combined with the customized apartment layout adaptation specifications, a two-dimensional customization basis is formed, and a finished apartment layout plan is customized accordingly. The experience result generation module is used by users to experience the customized finished apartment plan as a physical sample, judge the living suitability, and obtain the experience result. The iterative modification module is used to iteratively modify the customized apartment layout adaptation specifications based on the experience results until the user's needs are met.
[0017] As can be seen from the above technical solution, the present invention provides an automated derivation method for customized apartment layouts for end users. Compared with the prior art, the present invention has the following advantages: 1. This invention quantifies the user's desired apartment layout and the actual apartment layout, accurately identifies the user's needs, eliminates the designer's subjective judgment, and makes large-scale customization more efficient.
[0018] 2. This invention trains the model by inputting sample residential demand data, corresponding suitable apartment sizes and layout types from the training set into the model, continuously optimizing the model parameters until the preset adaptation accuracy requirements are met, and finally constructing an apartment adaptation recommendation algorithm model that includes a size recommendation model and a layout type recommendation model. The two models are modeled independently, and the size and layout adaptation accuracy are improved simultaneously. The customization efficiency is high, the recommendation standard is unified, and multi-dimensional needs are fully covered, making personalized adaptation more comprehensive and detailed.
[0019] 3. This invention uses a weighted summation formula for the final apartment type parameters to similarly weight and fuse the basic layout type and target layout type of the customized apartment type to determine the final apartment type layout. By integrating trial living experience and original needs, the final solution is more balanced and more in line with actual living conditions. Attached Figure Description
[0020] Figure 1 This is a flowchart illustrating an automated derivation method for customized apartment layouts for end users, as described in this invention.
[0021] Figure 2 This is a schematic diagram of the process S1 in this invention.
[0022] Figure 3 This is a schematic diagram of the process of S13 in this invention.
[0023] Figure 4 This is a schematic diagram of the process S5 in this invention.
[0024] Figure 5 This is a system block diagram of an automated inference system for customized apartment layouts for end users, as described in this invention. Detailed Implementation
[0025] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.
[0026] like Figure 1 As shown in the figure, an automated derivation method for customized apartment layouts for end users, according to this embodiment, includes the following steps: S1: Establish a housing type adaptation recommendation algorithm model; S2: Collect users' raw spoken language data in layers and convert the raw spoken language data into housing demand data, which includes basic housing needs, functional preference needs and personalized customization needs; S3: Input residential demand data into the apartment type adaptation recommendation algorithm model to generate a virtual trial apartment template and customized apartment type adaptation specifications for users. S4: Organize users to experience the virtual trial living room model through a three-dimensional virtual space, and mark the trial living feedback label of the model according to the experience. The trial living feedback label includes the fit, the optimization, and the rejection. S5: Adjust the adaptation specifications of the apartment template based on the trial stay feedback tags, and combine them with the customized apartment adaptation specifications to form a two-dimensional customization basis, and customize the finished apartment plan accordingly. S6: Users experience the customized finished apartment plan in person to determine its suitability for living and obtain the experience results; S7: Iteratively modify the customized apartment layout specifications based on user experience results until they meet user needs.
[0027] like Figure 2 As shown, S1 includes: S11: Classify and enter the basic attributes of the apartment type and the corresponding layout attributes into the database for archiving, collect complete living needs details of a large number of sample households, and form a sample living needs dataset; S12: Organize sample residents to experience various apartment layout options with different sizes and record their feedback on the size and layout of the apartments. S13: Calculate the redundancy parameters of the apartment layout based on sample residential demand data, suitable apartment size and apartment layout type; S14: Use the sample housing demand data as experience samples, divide them into training set and test set according to a preset ratio, and build a deep learning neural network algorithm model. S15: Input the sample housing demand data, corresponding suitable apartment size and apartment layout type into the model for training, continuously optimize the model parameters until the preset adaptation accuracy requirements are met, and finally construct an apartment adaptation recommendation algorithm model that includes a size recommendation model and a layout type recommendation model. The formula for calculating the redundancy parameter of the apartment layout is as follows:
[0028] in, For the first The first type of apartment layout Redundancy parameters in apartment layout corresponding to various apartment sizes. For the first The first type of apartment layout Average standard living requirements corresponding to various apartment sizes For the first The first type of apartment layout Actual apartment layout data corresponding to various apartment sizes. The number of sample households For the first The first type of apartment layout The first type of apartment size corresponds to the Housing needs data for a sample of households.
[0029] In practical applications, taking a family of three as an example, the core customization requirements are a three-bedroom apartment of about 90 square meters, separation of living and sleeping areas, a separate play area for children, and ample natural light in the living room. The goal is to adapt to the personalized customization of basic housing needs, and the implementation city is a second-tier city. Specifically, based on samples of 90㎡ three-bedroom apartments and layouts with separation of living and sleeping areas, the average lighting requirement of the living room for this type of apartment was calculated to be μ=0.45. The actual lighting coefficient of the living room in the building layout was L=0.4. Through the redundant parameter formula, σ=|0.45-0.4|×√[1 / 50×Σ(sample lighting requirement-0.45)²]=0.03 was calculated. The deviation between the quantitative lighting requirement and the actual layout was clearly 0.03, rather than a vague statement of insufficient lighting.
[0030] like Figure 3 As shown, S13 includes: S131: Classify the sample residential demand data according to the type of apartment layout that is suitable for the apartment type, and further subdivide the same layout type according to the size of the apartment type. S132: Calculate the mean of sample living demand data corresponding to each type of apartment size, and perform linear fitting on the mean to obtain the fitted data; S133: Based on the fitted data, derive the standard living requirements data corresponding to different apartment sizes under each apartment layout type; extract the actual apartment layout data corresponding to each apartment layout type and each apartment size; S134: Calculate the sample dispersion of the difference between standard residential demand data and actual unit layout data, and calculate the redundancy parameter of unit layout. The formula for calculating the mean of the sample housing demand data is as follows:
[0031] in, For the first The demand weights of each sample household.
[0032] like Figure 4 As shown, adjusting the floor plan specifications based on trial stay feedback tags includes the following steps: S51: Determine whether the space lighting requirements in the user's residential needs data are within the preset lighting requirements threshold range, and adjust the size and layout type of the apartment model according to the lighting adjustment coefficient; S52: Determine whether the functional area width requirement in the user's residential needs data is within the preset functional area width threshold range, and adjust the size and layout type of the apartment template accordingly through the functional area width adjustment coefficient; S53: Determine whether the user's residential needs data on circulation smoothness is within the preset circulation smoothness threshold range, and adjust the size and layout type of the apartment model according to the circulation optimization coefficient; S54: Determine whether the living area requirement in the user's living demand data is within the preset area requirement threshold range, and adjust the size and layout type of the apartment template accordingly through the area adaptation coefficient; S55: Based on the results of four adjustments, determine the size and layout type of the apartment model that precisely matches the user's living needs.
[0033] Furthermore, the process of creating a customized finished apartment plan based on the adjusted apartment template specifications and the customized apartment specifications obtained from user living needs data includes the following steps: Based on the preset mapping rules between customized apartment types and apartment templates, the adjusted apartment template dimensions and layout types are converted into the corresponding customized apartment basic dimensions and basic layout types. The target size and layout type of the customized apartment are directly derived from the user's living needs data; the base size and target size of the customized apartment are weighted and summed using the final apartment parameter weighted summation formula to obtain the final apartment size; Similarly, the weighted summation formula for the final apartment type parameters is used to weight and fuse the basic layout type of the customized apartment type with the target layout type to determine the final apartment type. Based on the final apartment dimensions and layout type, complete the design and production of the customized finished apartment plan; The final weighted summation formula for the apartment layout parameters is as follows:
[0034] in, For the final apartment layout parameters, Based on the weights of the basic parameters, To customize the basic parameters of the apartment layout, To customize the target parameters for the apartment type.
[0035] Furthermore, establishing a unit type adaptation recommendation algorithm model that includes a size recommendation model and a layout type recommendation model includes the following steps: First, build a size recommendation model independently, then build a layout type recommendation model separately. After both models meet the preset adaptation accuracy requirements, integrate and optimize them to form a complete apartment type adaptation recommendation algorithm model.
[0036] Furthermore, establishing a size recommendation model includes the following steps: Extract the redundant parameters of the apartment layout corresponding to the sample housing demand data, calculate the difference between the sample housing demand data and the corresponding redundant parameters, and obtain the corrected new housing demand data. Interpolation is used to match the area requirements in the new housing demand data with the basic area in the apartment layout data to obtain the preliminary apartment dimensions corresponding to the area requirements. By using matrix similarity analysis, the spatial dimension data in the new housing demand data is compared with the functional area dimension data in the apartment layout data to obtain the reference apartment dimensions corresponding to the spatial dimensions; The sample housing demand data and corresponding suitable apartment sizes in the training set are input into the established deep learning neural network algorithm model for training. The model parameters are continuously adjusted until the first preset adaptation accuracy is met, and the apartment sizes derived by the model are obtained. Using a linear regression algorithm, the preliminary apartment size, reference apartment size, model-derived apartment size, and sample-fitted apartment size are integrated. The deep learning neural network algorithm model is further optimized until the second preset fitting accuracy is met, and finally a size recommendation model is formed.
[0037] Furthermore, establishing a layout type recommendation model includes the following steps: Extract the redundant parameters of the apartment layout corresponding to the sample housing demand data, calculate the difference between the sample housing demand data and the corresponding redundant parameters, and obtain the corrected new housing demand data. By using matrix similarity analysis, the functional zoning demand data in the new housing demand data is compared with the zoning layout data in the apartment layout data to obtain the preliminary apartment layout type corresponding to the functional zoning. By using matrix similarity analysis, the circulation design requirements data in the new housing demand data are compared with the circulation planning data in the apartment layout data to obtain the reference apartment layout type corresponding to the circulation design. The sample residential demand data and corresponding suitable apartment layout types in the training set are input into the established deep learning neural network algorithm model for training. The model parameters are continuously adjusted until the first preset adaptation accuracy is met, and the apartment layout type derived by the model is obtained. Using a linear regression algorithm, the preliminary apartment layout types, reference apartment layout types, model-derived apartment layout types, and sample-fitted apartment layout types are integrated. The deep learning neural network algorithm model is further optimized until the second preset fitting accuracy is met, and finally, a layout type recommendation model is formed.
[0038] Specifically, the homeowner's original requirement (living room daylight factor 0.5) is reduced by a redundant parameter of 0.03, resulting in a revised value of 0.47. This avoids recommending a solution that is "theoretically 0.5 daylight factor but cannot be achieved in the building," making the model input more closely match the actual implementation conditions. Size recommendation model: Focusing on numerical needs, it accurately recommends "living room 25㎡ (instead of 28㎡, with space reserved for play area), children's play area 8㎡, master bedroom 14㎡, secondary bedroom 10㎡, kitchen and bathroom 13㎡", with size deviation controlled within ±0.5㎡; Layout type recommendation model: Focusing on structural needs, the recommended layout is "South-facing living room (matching lighting requirements) + play area adjacent to the living room but separated by a low cabinet (achieving separation of active and quiet areas + accessible to parents) + kitchen and bathroom in a centralized layout (shortest traffic flow)". The layout adaptability is 60% higher than that of manual layout. Using weighted formula Because the homeowners provided detailed feedback during their trial stay (noting specific locations of narrow passageways and risks of bumping into children), α was set to 0.5 (basic parameter weight), and the calculation was as follows: The final play area area = 0.5 × 7 + 0.5 × 8 = 7.5㎡; Final traffic flow width = 0.5 × 1.2 + 0.5 × 1 = 1.1m; The design satisfies both the requirement that "the play area is close to the owner's needs" and the safety issue of "narrow circulation routes," balancing theoretical requirements with actual living experience. If the homeowner only stays briefly (without specifying any issues), α can be adjusted to 0.4, prioritizing the original requirements; if the homeowner stays multiple times and provides detailed feedback, α should be adjusted to 0.6, prioritizing the experience and adapting to different customization stages. Through the above three core steps, a 90㎡ three-bedroom apartment was finally customized for the homeowner: Dimensions: Living room 25㎡ (lighting factor 0.47), children's play area 7.5㎡, master bedroom 14㎡, second bedroom 10㎡, kitchen and bathroom 13㎡; Layout: South-facing living room + play area adjacent to the living room (separated by a low cabinet) + 1.1m wide circulation path; After the homeowners tried living in the physical model, they reported that "the lighting is sufficient, the children's play area is the right size and the traffic flow is safe, and the separation of active and quiet areas meets expectations." The customized plan was approved on the first try, with no rework, which improved efficiency by 80% and adaptation accuracy by 70% compared to traditional manual customization.
[0039] Furthermore, converting raw spoken data into residential demand data includes: Establish a mapping library between common customer language and professional apartment terminology. The mapping library contains common customer colloquial needs and their corresponding unique professional expressions. The collected raw spoken data from users is matched sentence by sentence with common expressions in the mapping library to initially generate corresponding professional requirement expressions; Professional designers were assigned to manually verify the initial converted professional requirements descriptions, and, based on user supplementary explanations, correct any deviations in the descriptions and supplement any missing details. The verified professional requirements are quantified into professional data that can be entered into the model, and then integrated with the basic requirements data filled in on paper to form professional residential requirements data.
[0040] like Figure 5 As shown, an automated apartment layout customization system for end users includes: The model building module is used to build a model for apartment type matching recommendation algorithms. The data acquisition and processing module is used to collect users' raw spoken language data in layers and convert the raw spoken language data into residential demand data; The virtual template generation module is used to input residential demand data into the apartment type adaptation recommendation algorithm model to generate a virtual trial apartment template and customized apartment type adaptation specifications for the user. The tag generation module is used to organize users to experience virtual trial living room models in a 3D virtual space and to label the trial living room models with trial living feedback tags based on their experience. The apartment layout customization module is used to adjust the adaptation specifications of the apartment layout template based on the trial stay feedback tags. Combined with the customized apartment layout adaptation specifications, a two-dimensional customization basis is formed, and a finished apartment layout plan is customized accordingly. The experience result generation module is used by users to experience the customized finished apartment plan as a physical model, judge the living suitability, and obtain the experience result. The iterative modification module is used to iteratively modify the customized apartment layout specifications based on user experience results until they meet user needs.
[0041] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this application is generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)).
[0042] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.
[0043] The various embodiments in this specification are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions of the method embodiments.
[0044] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A terminal-oriented automated derivation method for customized apartment layouts, characterized in that, Includes the following steps: S1: Establish a housing type adaptation recommendation algorithm model; S2: Collect users' raw spoken language data in layers and convert the raw spoken language data into housing demand data, which includes basic housing demand, functional preference demand and personalized customization demand; S3: Substitute the residential demand data into the apartment type adaptation recommendation algorithm model to generate a virtual trial apartment template and customized apartment type adaptation specifications for the user. S4: Organize users to experience the virtual trial living room model through a three-dimensional virtual space, and mark the trial living feedback label of the model according to the experience. The trial living feedback label includes the adaptation item, the optimization item, and the rejection item. S5: Adjust the adaptation specifications of the apartment template based on the trial stay feedback tags, and combine them with the customized apartment adaptation specifications to form a two-dimensional customization basis, and customize the finished apartment plan accordingly. S6: The user experiences the customized finished apartment plan as a physical sample, judges its suitability for living, and obtains the experience results; S7: Iteratively modify the customized apartment layout specifications based on the experience results until the user's needs are met.
2. The automated derivation method for customized apartment layouts for end-users according to claim 1, characterized in that: S1 includes: S11: Classify and enter the basic attributes of the apartment type and the corresponding layout attributes into the database for archiving, collect complete living needs details of a large number of sample households, and form a sample living needs dataset; S12: Organize sample residents to experience various apartment layout options with different sizes and record their feedback on the size and layout of the apartments. S13: Calculate the redundancy parameters of the apartment layout based on sample residential demand data, suitable apartment size and apartment layout type; S14: Use the sample housing demand data as experience samples, divide them into training set and test set according to a preset ratio, and build a deep learning neural network algorithm model. S15: Input the sample housing demand data, corresponding suitable apartment size and apartment layout type into the model for training, continuously optimize the model parameters until the preset adaptation accuracy requirements are met, and finally construct an apartment adaptation recommendation algorithm model that includes a size recommendation model and a layout type recommendation model. The formula for calculating the redundancy parameter of the apartment layout is as follows: in, For the first The first type of apartment layout Redundancy parameters in apartment layout corresponding to various apartment sizes. For the first The first type of apartment layout Average standard living requirements corresponding to various apartment sizes For the first The first type of apartment layout Actual apartment layout data corresponding to various apartment sizes. The number of sample households For the first The first type of apartment layout The first type of apartment size corresponds to the Housing needs data for a sample of households.
3. The automated derivation method for customized apartment layouts for end-users according to claim 2, characterized in that: S13 includes: S131: Classify the sample residential demand data according to the type of apartment layout that is suitable for the apartment type, and further subdivide the same layout type according to the size of the apartment type. S132: Calculate the mean of sample living demand data corresponding to each type of apartment size, and perform linear fitting on the mean to obtain the fitted data; S133: Based on the fitted data, derive the standard living requirements data corresponding to different apartment sizes under each apartment layout type; extract the actual apartment layout data corresponding to each apartment layout type and each apartment size; S134: Calculate the sample dispersion of the difference between standard residential demand data and actual unit layout data, and calculate the redundancy parameter of unit layout. The formula for calculating the mean of the sample housing demand data is as follows: in, For the first The demand weights of each sample household.
4. The automated derivation method for terminal-oriented apartment layout customization according to claim 3, characterized in that: The process of adjusting the apartment layout specifications based on trial stay feedback tags includes the following steps: S51: Determine whether the space lighting requirements in the user's residential needs data are within the preset lighting requirements threshold range, and adjust the size and layout type of the apartment model according to the lighting adjustment coefficient; S52: Determine whether the functional area width requirement in the user's residential needs data is within the preset functional area width threshold range, and adjust the size and layout type of the apartment template accordingly through the functional area width adjustment coefficient; S53: Determine whether the user's residential needs data on circulation smoothness is within the preset circulation smoothness threshold range, and adjust the size and layout type of the apartment model according to the circulation optimization coefficient; S54: Determine whether the living area requirement in the user's living demand data is within the preset area requirement threshold range, and adjust the size and layout type of the apartment template accordingly through the area adaptation coefficient; S55: Based on the results of four adjustments, determine the size and layout type of the apartment model that precisely matches the user's living needs.
5. The automated derivation method for terminal-oriented apartment layout customization according to claim 4, characterized in that: The process of creating a customized finished apartment plan based on the adjusted apartment template specifications and the customized apartment specifications obtained from user living needs data includes the following steps: Based on the preset mapping rules between customized apartment types and apartment templates, the adjusted apartment template dimensions and layout types are converted into the corresponding customized apartment basic dimensions and basic layout types. The target size and layout type of the customized apartment are directly derived from the user's living needs data; the base size and target size of the customized apartment are weighted and summed using the final apartment parameter weighted summation formula to obtain the final apartment size; Similarly, the weighted summation formula for the final apartment type parameters is used to weight and fuse the basic layout type of the customized apartment type with the target layout type to determine the final apartment type. Based on the final apartment dimensions and layout type, complete the design and production of the customized finished apartment plan; The weighted summation formula for the final apartment layout parameters is as follows: in, For the final apartment layout parameters, Based on the weights of the basic parameters, To customize the basic parameters of the apartment layout, To customize the target parameters for the apartment type.
6. The automated derivation method for customized apartment layouts for terminal devices according to claim 5, characterized in that: The weighted summation formula for the final apartment layout parameters is as follows: in, For the final apartment layout parameters, Based on the weights of the basic parameters, To customize the basic parameters of the apartment layout, To customize the target parameters for the apartment type.
7. The automated derivation method for customized apartment layouts for end-users according to claim 6, characterized in that: The process of establishing a unit type adaptation recommendation algorithm model, which includes a size recommendation model and a layout type recommendation model, includes the following steps: First, build a size recommendation model independently, then build a layout type recommendation model separately. After both models meet the preset adaptation accuracy requirements, integrate and optimize them to form a complete apartment type adaptation recommendation algorithm model.
8. The automated derivation method for customized apartment layouts for end-users according to claim 7, characterized in that: The establishment of the layout type recommendation model includes the following steps: Extract the redundant parameters of the apartment layout corresponding to the sample housing demand data, calculate the difference between the sample housing demand data and the corresponding redundant parameters, and obtain the corrected new housing demand data. By using matrix similarity analysis, the functional zoning demand data in the new housing demand data is compared with the zoning layout data in the apartment layout data to obtain the preliminary apartment layout type corresponding to the functional zoning. By using matrix similarity analysis, the circulation design requirements data in the new housing demand data are compared with the circulation planning data in the apartment layout data to obtain the reference apartment layout type corresponding to the circulation design. The sample residential demand data and corresponding suitable apartment layout types in the training set are input into the established deep learning neural network algorithm model for training. The model parameters are continuously adjusted until the first preset adaptation accuracy is met, and the apartment layout type derived by the model is obtained. Using a linear regression algorithm, the preliminary apartment layout types, reference apartment layout types, model-derived apartment layout types, and sample-fitted apartment layout types are integrated. The deep learning neural network algorithm model is further optimized until the second preset fitting accuracy is met, and finally, a layout type recommendation model is formed.
9. The automated derivation method for customized apartment layouts for end-users according to claim 8, characterized in that: The process of converting raw spoken data into residential demand data includes: Establish a mapping library between common customer language and professional apartment terminology. The mapping library contains common customer colloquial needs and their corresponding unique professional expressions. The collected raw spoken data from users is matched sentence by sentence with common expressions in the mapping library to initially generate corresponding professional requirement expressions; Professional designers were assigned to manually verify the initial converted professional requirements descriptions, and, based on user supplementary explanations, correct any deviations in the descriptions and supplement any missing details. The verified professional requirements are quantified into professional data that can be entered into the model, and then integrated with the basic requirements data filled in on paper to form professional residential requirements data.
10. A terminal-oriented automated derivation system for customized apartment layouts, applied to the terminal-oriented automated derivation method for customized apartment layouts as described in any one of claims 1-9, characterized in that, include: The model building module is used to build a model for apartment type matching recommendation algorithms. The data acquisition and processing module is used to collect users' raw spoken language data in layers and convert the raw spoken language data into residential demand data; The virtual template generation module is used to input the residential demand data into the apartment type adaptation recommendation algorithm model to generate a virtual trial apartment type template and customized apartment type adaptation specifications for the user. The tag generation module is used to organize users to experience virtual trial living room models in a 3D virtual space and to label the trial living room models with trial living feedback tags based on their experience. The apartment layout customization module is used to adjust the adaptation specifications of the apartment layout template based on the trial stay feedback tags. Combined with the customized apartment layout adaptation specifications, a two-dimensional customization basis is formed, and a finished apartment layout plan is customized accordingly. The experience result generation module is used by users to experience the customized finished apartment plan as a physical sample, judge the living suitability, and obtain the experience result. The iterative modification module is used to iteratively modify the customized apartment layout adaptation specifications based on the experience results until the user's needs are met.