Building design system based on intelligent optimization

Abstract

The invention discloses a building design system based on intelligent optimization, and the system collects user design interaction data through a building design platform, carries out the statistics of design regions, components, time and frequency, and forms user design interaction features. A PCA principal component analysis method is adopted to extract principal components of user interaction features, and a K-means algorithm is utilized to form a plurality of design user groups in a clustering mode. And in a preset time period, evaluating the real-time resource loading capability of the design user group, optimizing the pre-loading scheme in combination with resource loading capability information, and generating a model loading scheme based on intelligent optimization. According to the method, the comprehensive design efficiency of large-scale and different professional design users can be improved, the intelligence and automation of the platform are improved, the design period is shortened, and the collaborative design efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the field of architectural design, and more specifically, to an architectural design system based on intelligent optimization. Background Art

[0002] With the rapid development of information technology and the digital transformation of the architectural design industry, traditional architectural design methods have been unable to meet the current needs of efficient, personalized and sustainable development. The architectural design process involves a large amount of data, including building size, material selection, structural layout, energy consumption simulation, etc. The management and optimization of these data have become the key to improving design efficiency and quality.

[0003] However, most existing architectural design software remains at the level of auxiliary drawing and simple calculation, lacking intelligent data analysis and optimization capabilities, resulting in long design cycles, high costs, and insufficient innovation. In traditional design platforms, the data model loading process is often slow and inefficient. For large-scale design systems, the file preloading solution is often just a simple analysis of user design preferences, lacking targeted optimization analysis, and difficult to meet modern design needs. Summary of the invention

[0004] The present invention overcomes the defects of the prior art and proposes a building design system based on intelligent optimization.

[0005] A first aspect of the present invention provides a building design method based on intelligent optimization, comprising: Through the architectural design platform, user design interaction data is obtained. Through the design interaction data, the design areas, design components, design time, and design frequency of different users are counted to obtain user design interaction characteristics; Based on the user design interaction features, the principal component selection of the user interaction features is performed through the PCA principal component analysis method to generate the principal component features of each user; Based on the K-means algorithm, the principal component features of each user are clustered to form multiple designed user groups; In a preset historical time period, network interaction data and server preloading information of a design user group are collected, and real-time resource loading capability evaluation of the design user group is performed based on the network interaction data and server preloading information to generate resource loading capability information; The target building model is divided into multiple design areas. Within the preset historical time period, the real-time design interaction instructions of a design user group are collected. The design areas browsed, designed, and edited by users are marked and counted through the interaction instructions, and the key areas and non-key areas are screened out. Combined with the resource loading capacity information, the pre-loading plan for the design user group is optimized to generate a model loading plan based on intelligent optimization.

[0006] In this solution, the user design interaction data is obtained through the architectural design platform, and the design areas, design components, design time, and design frequency of different users are counted through the design interaction data to obtain the user design interaction characteristics, which are specifically: Build a cloud computing-based architectural design platform; Data transmission between multiple user terminals through the building design platform; Based on the network transmission between the architectural design platform and multiple user terminals, the user design interaction data is obtained; According to the design areas, design components, design time and design frequency of different users, data statistics and initial user characteristics analysis are performed on the user design interaction data to obtain the user design interaction characteristics.

[0007] In this solution, the user interaction features are designed based on the user, and the principal component selection is performed on the user interaction features through the PCA principal component analysis method to generate the principal component features of each user, specifically: Standardize the data of user design interaction features; Based on the PCA analysis method, the characteristic variables of user design interaction features are extracted and the covariance matrix is ​​constructed; Perform eigenvalue decomposition on the covariance matrix to form eigenvalues ​​and corresponding eigenvectors; Sorting is performed based on eigenvalues, and the first N eigenvectors are selected as principal components according to preset standards; A new feature space is constructed based on the selected principal component feature vectors. Through the new feature space, data of user interaction features is reconstructed and principal component features are generated.

[0008] In this solution, the principal component features of each user are clustered based on the K-means algorithm to form multiple design user groups, specifically: Construct a clustering model based on K-means and set K center points to initialize the clustering model; Import the principal component features of each user into the clustering model, use the principal component features as clustering sample data, use each user as a clustering unit, calculate the distance from the principal component features to the center point, and group the sample data based on the distance; Repeatedly calculate the new center point and compare the new center point with the original center point until the center point no longer moves. At this time, record the current clustering result; According to the current clustering results, users are clustered and grouped to form multiple designed user groups.

[0009] In this solution, the network interaction data and server preloading information of a design user group are collected within a preset historical time period, and the real-time resource loading capability of the design user group is evaluated through the network interaction data and server preloading information to generate resource loading capability information, specifically: In a preset historical time period, a design user group is used as an analysis unit to collect network interaction data and server preloading information of a design user group; The network interaction data includes port information, network status, and data transmission rate information; The server preloading information includes historical loading data volume, loading waiting time, and preloading success rate information; Through network interaction data and server preloading information, a design user group is evaluated based on network latency, network rate, and average loaded data volume, and resource loading capacity information is generated.

[0010] In this solution, the target building model is divided into multiple design areas, and real-time design interaction instructions of a design user group are collected within a preset historical time period. The design areas browsed, designed, and edited by users are marked and counted through interaction instructions, and key areas and non-key areas are screened out. Combined with resource loading capacity information, the pre-loading solution of the design user group is optimized to generate a model loading solution based on intelligent optimization, specifically: Through the architectural design platform, the target building model is divided into multiple design areas; In a preset historical time period, the real-time design interaction instructions of all users in a design user group are collected, and the interaction instructions are identified. Through instruction identification, the design areas corresponding to the user's browsing, design, and editing behaviors are marked and counted, and the interaction frequency of different design areas is analyzed to mark high-frequency interaction areas and low-frequency interaction areas; The high-frequency interaction areas and low-frequency interaction areas are marked as key areas and non-key areas respectively.

[0011] In this solution, the preloading solution is optimized for the design user group in combination with the resource loading capability information to generate a model loading solution based on intelligent optimization, and also includes: In a preset historical time period, real-time design interaction instructions of a design user group are collected, instruction recognition and user behavior analysis are performed on the interaction instructions, the instructions are associated with the user behavior, the instructions are sorted based on the time dimension of the collected interaction instructions, and a user behavior sequence is generated based on the instruction sorting result; In a preset historical time period, based on whether the design interface in which the user is located is a key area or a non-key area during the interactive design, the user behavior sequence is divided to generate a first behavior sequence and a second behavior sequence; Build a prediction model based on LSTM and import historical user behavior sequence data into the prediction model for model training; A prediction cycle is set, and the first behavior sequence and the second behavior sequence are respectively imported into the prediction model to predict user behavior, and a first predicted behavior and a second predicted behavior are obtained; According to the first predicted behavior and the second predicted behavior, the design behavior pattern of the design user group is analyzed, and the loading plan analysis is performed on the key areas and non-key areas. Combined with the resource loading capability information, the loading model file, loading file type, and file loading amount parameters are set, the original loading plan is optimized, and a model loading plan for the design user group is generated.

[0012] The second aspect of the present invention further provides an architectural design system based on intelligent optimization, the system comprising: a memory, a processor, the memory comprising an architectural design program based on intelligent optimization, the architectural design program based on intelligent optimization being executed by the processor to implement the following steps: Through the architectural design platform, user design interaction data is obtained. Through the design interaction data, the design areas, design components, design time, and design frequency of different users are counted to obtain user design interaction characteristics; Based on the user design interaction features, the principal component selection of the user interaction features is performed through the PCA principal component analysis method to generate the principal component features of each user; Based on the K-means algorithm, the principal component features of each user are clustered to form multiple designed user groups; In a preset historical time period, network interaction data and server preloading information of a design user group are collected, and real-time resource loading capability evaluation of the design user group is performed based on the network interaction data and server preloading information to generate resource loading capability information; The target building model is divided into multiple design areas. Within the preset historical time period, the real-time design interaction instructions of a design user group are collected. The design areas browsed, designed, and edited by users are marked and counted through the interaction instructions, and the key areas and non-key areas are screened out. Combined with the resource loading capacity information, the pre-loading plan for the design user group is optimized to generate a model loading plan based on intelligent optimization.

[0013] The third aspect of the present invention also provides a computer-readable storage medium, which includes an architectural design program based on intelligent optimization. When the architectural design program based on intelligent optimization is executed by a processor, the steps of the architectural design method based on intelligent optimization as described in any one of the above items are implemented.

[0014] The present invention discloses an architectural design system based on intelligent optimization. The present invention collects user design interaction data through an architectural design platform, and counts the design area, component, time and frequency to form user design interaction characteristics. The PCA principal component analysis method is used to extract the principal components of the user interaction characteristics, and the K-means algorithm is used for clustering to form multiple design user groups. Within a preset time period, the real-time resource loading capability of the design user group is evaluated, and the preloading scheme is optimized in combination with the resource loading capability information to generate a model loading scheme based on intelligent optimization. Through the present invention, it is possible to improve the comprehensive design efficiency of large-scale and different professional design users, enhance the intelligence and automation of the platform, shorten the design cycle, and improve the efficiency of collaborative design. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 A flow chart of a building design method based on intelligent optimization of the present invention is shown; Figure 2 A block diagram of an architectural design system based on intelligent optimization according to the present invention is shown. DETAILED DESCRIPTION

[0016] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0017] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0018] Figure 1 A flow chart of a building design method based on intelligent optimization of the present invention is shown.

[0019] like Figure 1 As shown, the first aspect of the present invention provides a building design method based on intelligent optimization, comprising: S102, obtaining user design interaction data through the architectural design platform, and performing statistics on the design areas, design components, design time, and design frequency of different users through the design interaction data to obtain user design interaction characteristics; S104, based on the user design interaction features, principal component selection is performed on the user interaction features by using a PCA principal component analysis method to generate principal component features for each user; S106, clustering the principal component features of each user based on a K-means algorithm to form multiple designed user groups; S108, collecting network interaction data and server preloading information of a design user group within a preset historical time period, performing real-time resource loading capability evaluation on the design user group based on the network interaction data and server preloading information, and generating resource loading capability information; S110, divide the target building model into multiple design areas, collect real-time design interaction instructions of a design user group within a preset historical time period, mark and count the design areas browsed, designed, and edited by users through interaction instructions, and screen out key areas and non-key areas. Combined with resource loading capacity information, optimize the preloading plan for the design user group and generate a model loading plan based on intelligent optimization.

[0020] According to an embodiment of the present invention, the user design interaction data is obtained through the architectural design platform, and the design areas, design components, design time, and design frequency of different users are counted through the design interaction data to obtain user design interaction features, which are specifically: Build a cloud computing-based architectural design platform; Data transmission between multiple user terminals through the building design platform; Based on the network transmission between the architectural design platform and multiple user terminals, the user design interaction data is obtained; According to the design areas, design components, design time and design frequency of different users, data statistics and initial user characteristics analysis are performed on the user design interaction data to obtain the user design interaction characteristics.

[0021] It should be noted that the user design interaction features are user interaction features of the initial primary analysis, including feature data of higher dimensions, high data redundancy, and more variable features. The user terminal is the client.

[0022] According to an embodiment of the present invention, the user interaction features are designed based on the user, and the principal component selection is performed on the user interaction features through the PCA principal component analysis method to generate the principal component features of each user, specifically: Standardize the data of user design interaction features; Based on the PCA analysis method, the characteristic variables of user design interaction features are extracted and the covariance matrix is ​​constructed; Perform eigenvalue decomposition on the covariance matrix to form eigenvalues ​​and corresponding eigenvectors; Sorting is performed based on eigenvalues, and the first N eigenvectors are selected as principal components according to preset standards; A new feature space is constructed based on the selected principal component feature vectors. Through the new feature space, the user interaction features are reconstructed and principal component features are generated.

[0023] It should be noted that the covariance matrix reflects the strength and direction of the linear relationship between variables. Through the covariance matrix, the data can be reduced in dimension in the future. The principal component feature is the feature data formed by reducing the data dimension based on the user-designed interaction features, which is used for the subsequent feature analysis and clustering of user interaction behaviors.

[0024] It is worth mentioning here that in a large-scale design platform, there are users with a large amount of data who design, edit, and update architectural models at the same time. The platform data transmission volume is large during the process, and the user interaction data is relatively complex, including component selection, design element editing, model viewing, model comparison, etc. in the design process, and users of different design types correspond to different design contents, such as users responsible for different professional models. Therefore, the collected interaction data has high-dimensional behavioral characteristics. If a simple behavioral model analysis is performed on users based on traditional methods, it will lead to large individual differences and it will be difficult to classify users by behavioral characteristics. Based on this, the present invention performs data dimensionality reduction analysis on user interaction data through the PCA principal component analysis method, and further screens out the principal component characteristics for subsequent user clustering and grouping.

[0025] According to an embodiment of the present invention, the principal component features of each user are clustered based on the K-means algorithm to form multiple designed user groups, specifically: Construct a clustering model based on K-means and set K center points to initialize the clustering model; Import the principal component features of each user into the clustering model, use the principal component features as clustering sample data, use each user as a clustering unit, calculate the distance from the principal component features to the center point, and group the sample data based on the distance; Repeatedly calculate the new center point and compare the new center point with the original center point until the center point no longer moves. At this time, record the current clustering result; According to the current clustering results, users are clustered and grouped to form multiple designed user groups.

[0026] It should be noted that the clustering result includes the grouping of the principal component features, and corresponds to the grouping of the users. The user groups are designed to be grouped based on the user interaction behavior features, and the interaction features of the users in each group have certain similarities.

[0027] According to an embodiment of the present invention, within a preset historical time period, network interaction data and server preloading information of a design user group are collected, and real-time resource loading capability evaluation of the design user group is performed through the network interaction data and the server preloading information to generate resource loading capability information, specifically: In a preset historical time period, a design user group is used as an analysis unit to collect network interaction data and server preloading information of a design user group; The network interaction data includes port information, network status, and data transmission rate information; The server preloading information includes historical loading data volume, loading waiting time, and preloading success rate information; Through network interaction data and server preloading information, a design user group is evaluated based on network latency, network rate, and average loaded data volume, and resource loading capacity information is generated.

[0028] It should be noted that the preloading success rate information, i.e., the ratio of the amount of successfully preloaded data to the total load, reflects the preloading data capacity within a preset historical time period. Based on different user terminals and different network conditions, when users open, edit, and design building models, the preloading conditions of the models are different. The resource loading capacity information is used to reflect the average resource loading capacity of a design user group. Through this capacity information, targeted resource loading optimization solution analysis can be performed in the subsequent period.

[0029] According to an embodiment of the present invention, the target building model is divided into multiple design areas, and real-time design interaction instructions of a design user group are collected within a preset historical time period. The design areas browsed, designed, and edited by users are marked and counted through the interaction instructions, and key areas and non-key areas are screened out. Combined with the resource loading capability information, the preloading scheme of the design user group is optimized to generate a model loading scheme based on intelligent optimization, specifically: Through the architectural design platform, the target building model is divided into multiple design areas; In a preset historical time period, the real-time design interaction instructions of all users in a design user group are collected, and the interaction instructions are identified. Through instruction identification, the design areas corresponding to the user's browsing, design, and editing behaviors are marked and counted, and the interaction frequency of different design areas is analyzed to mark high-frequency interaction areas and low-frequency interaction areas; The high-frequency interaction areas and low-frequency interaction areas are marked as key areas and non-key areas respectively.

[0030] It should be noted that the design and editing focuses of different design users are in different design areas.

[0031] According to an embodiment of the present invention, the method of optimizing the preloading scheme for the design user group in combination with the resource loading capability information and generating a model loading scheme based on intelligent optimization further includes: In a preset historical time period, real-time design interaction instructions of a design user group are collected, instruction recognition and user behavior analysis are performed on the interaction instructions, the instructions are associated with the user behavior, the instructions are sorted based on the time dimension of the collected interaction instructions, and a user behavior sequence is generated based on the instruction sorting result; In a preset historical time period, based on whether the design interface in which the user is located is a key area or a non-key area during the interactive design, the user behavior sequence is divided to generate a first behavior sequence and a second behavior sequence; Build a prediction model based on LSTM and import historical user behavior sequence data into the prediction model for model training; A prediction cycle is set, and the first behavior sequence and the second behavior sequence are respectively imported into the prediction model to predict user behavior, and a first predicted behavior and a second predicted behavior are obtained; According to the first predicted behavior and the second predicted behavior, the design behavior pattern of the design user group is analyzed, and the loading plan analysis is performed on the key areas and non-key areas. Combined with the resource loading capability information, the loading model file, loading file type, and file loading amount parameters are set, the original loading plan is optimized, and a model loading plan for the design user group is generated.

[0032] It should be noted that the interactive instructions include client control instructions when the user browses, designs, and edits the corresponding design area. Through instruction recognition, pattern recognition of user behavior can be performed. In the present invention, by identifying and serializing instructions, a corresponding user behavior sequence can be generated, and user behavior prediction is performed based on a prediction model. The user behavior sequence corresponds one-to-one to the instruction sequence. The first behavior sequence corresponds to the behavior data generated when the user performs interactive design in the key area, and the second behavior sequence corresponds to the behavior data generated when the user performs interactive design in the non-key area. The first predicted behavior and the second predicted behavior are design prediction behaviors for the design user group in the key area and non-key area of ​​the model, and the user's subsequent preloading solution analysis.

[0033] It is worth mentioning that in the actual architectural design process, different users have different frequencies of editing, using, and updating content for different design areas, and different design focuses. For example, in the target architectural model, there are multiple professional models, and different users or users of different user groups have different key design areas for the model. There are also certain differences in the design areas of each client loading model, design model, and editing model. Therefore, in view of the differences between different users, the present invention analyzes the interactive characteristics and clusters the users, generates corresponding key and non-key design areas for each user group, and generates sequence data based on user design behavior based on the correlation between interactive instructions and behavior models. According to the sequence data, the LSTM prediction model in the present invention can be combined to evaluate user behavior, and the pre-loaded data can be adjusted to improve the comprehensive design efficiency of large-scale and different professional design users, improve the intelligence and automation of the platform, shorten the design cycle, and improve the efficiency of collaborative design. At the same time, the intelligent optimization of model loading is effectively carried out, so that each design user group has a different loading scheme to improve the overall design efficiency.

[0034] The key areas and non-key areas are used to determine whether the user group performs design interactions in the key high-frequency design model. The key areas are screened based on the user's interaction data over a period of time, specifically to the marked areas within the architectural design.

[0035] It is worth mentioning that traditional preloading solutions often simply analyze user design preferences and load commonly used tool models. However, for large-scale collaborative design and design platforms with multi-professional architectural models, traditional methods are inefficient and difficult to meet modern design needs.

[0036] According to an embodiment of the present invention, it also includes: Integrate the behavior data of a first predicted behavior and a second predicted behavior of a design user group to obtain predicted behavior data; The predicted behavior data is transformed into features and the features are analyzed based on the PCA principal component analysis method to generate a predicted principal component feature of the design user group; In the clustering model, the predicted principal component features are imported into the clustering model, and the distance between the predicted principal component features and the cluster center is calculated based on the standard Euclidean distance to determine whether the distance is greater than the threshold; If so, it is determined that there are users with behavioral patterns that deviate from the current design user group, the design interaction features of all users in the design user group are obtained, and the difference analysis of individual user behaviors is performed through the interaction features, and users whose behavioral features deviate from the design user group are screened out and marked as deviant users; The deviating user is deleted from the one design user group and the design user group is updated.

[0037] It should be noted that in the distance calculation with the cluster center, the cluster center selects the cluster center corresponding to the one designed user group, with a total of K center points, and each center point corresponds to a designed user group. The single user behavior difference analysis through interactive features is to perform deviation analysis on each user and the overall behavioral characteristics in the user design group, screen out users who deviate greatly from the overall user behavior characteristics in the group, and remove them from the group to form a new designed user group, thereby ensuring that each designed user group has users with similar behavioral characteristics, and achieving the pertinence and accuracy of subsequent loading plans. If so, it is judged that there are users whose behavior patterns deviate from the current designed user group, specifically judging the user behavior deviation by the distance between the predicted behavioral characteristics of the user group and the cluster center.

[0038] According to an embodiment of the present invention, it also includes: Based on multiple design user groups, the model synchronization time node of each design user group is obtained through the architectural design platform; By synchronizing the time nodes, the time interval between every two consecutive time nodes is counted, and the average time interval is calculated; In the LSTM prediction model, the length of the prediction cycle is set as the average time interval. After each data synchronization process, the user behavior deviation evaluation and user behavior anomaly evaluation are performed through the first prediction behavior and the second prediction behavior of a designed user group, and the data synchronization anomaly warning analysis is performed based on the evaluation results. After a prediction cycle, the client synchronization log data is obtained and sent to the architectural design platform.

[0039] It should be noted that in the actual design platform, due to the need for collaborative design, data synchronization is required. Each data synchronization requires real-time updating and sharing of the building model. However, due to the large-scale synchronization data operation and the server's synchronization data capabilities, synchronization anomalies and data consistency mismatches often occur. The abnormal state is often related to the user's abnormal design behavior. At the same time, it is highly concealed and difficult to mine and analyze.

[0040] Therefore, the present invention sets the prediction period and the data synchronization period to be consistent. After each data synchronization, it determines whether there are deviant users and abnormal user behaviors based on the predicted user behaviors. If there are abnormal behaviors, the synchronization nodes are marked, and the data synchronization is reviewed by obtaining the synchronization log, so as to reduce the probability of abnormalities in the synchronization data and realize abnormal warning.

[0041] Figure 2 A block diagram of an architectural design system based on intelligent optimization according to the present invention is shown.

[0042] The second aspect of the present invention further provides an architectural design system 2 based on intelligent optimization, the system comprising: a memory 21, a processor 22, the memory 21 comprising an architectural design program based on intelligent optimization, the architectural design program based on intelligent optimization being executed by the processor 22 to implement the following steps: Through the architectural design platform, user design interaction data is obtained. Through the design interaction data, the design areas, design components, design time, and design frequency of different users are counted to obtain user design interaction characteristics; Based on the user design interaction features, the principal component selection of the user interaction features is performed through the PCA principal component analysis method to generate the principal component features of each user; Based on the K-means algorithm, the principal component features of each user are clustered to form multiple designed user groups; In a preset historical time period, network interaction data and server preloading information of a design user group are collected, and real-time resource loading capability evaluation of the design user group is performed based on the network interaction data and server preloading information to generate resource loading capability information; The target building model is divided into multiple design areas. Within the preset historical time period, the real-time design interaction instructions of a design user group are collected. The design areas browsed, designed, and edited by users are marked and counted through the interaction instructions, and the key areas and non-key areas are screened out. Combined with the resource loading capacity information, the pre-loading plan for the design user group is optimized to generate a model loading plan based on intelligent optimization.

[0043] According to an embodiment of the present invention, the user design interaction data is obtained through the architectural design platform, and the design areas, design components, design time, and design frequency of different users are counted through the design interaction data to obtain user design interaction features, which are specifically: Build a cloud computing-based architectural design platform; Data transmission between multiple user terminals through the building design platform; Based on the network transmission between the architectural design platform and multiple user terminals, the user design interaction data is obtained; According to the design areas, design components, design time and design frequency of different users, data statistics and initial user characteristics analysis are performed on the user design interaction data to obtain the user design interaction characteristics.

[0044] It should be noted that the user design interaction features are user interaction features of the initial primary analysis, including feature data of higher dimensions, high data redundancy, and more variable features. The user terminal is the client.

[0045] According to an embodiment of the present invention, the user interaction features are designed based on the user, and the principal component selection is performed on the user interaction features through the PCA principal component analysis method to generate the principal component features of each user, specifically: Standardize the data of user design interaction features; Based on the PCA analysis method, the characteristic variables of user design interaction features are extracted and the covariance matrix is ​​constructed; Perform eigenvalue decomposition on the covariance matrix to form eigenvalues ​​and corresponding eigenvectors; Sorting is performed based on eigenvalues, and the first N eigenvectors are selected as principal components according to preset standards; A new feature space is constructed based on the selected principal component feature vectors. Through the new feature space, data of user interaction features is reconstructed and principal component features are generated.

[0046] It should be noted that the covariance matrix reflects the strength and direction of the linear relationship between variables. Through the covariance matrix, the data can be reduced in dimension in the future. The principal component feature is the feature data formed by reducing the data dimension based on the user-designed interaction features, which is used for the subsequent feature analysis and clustering of user interaction behaviors.

[0047] It is worth mentioning here that in a large-scale design platform, there are users with a large amount of data who design, edit, and update architectural models at the same time. The platform data transmission volume is large during the process, and the user interaction data is relatively complex, including component selection, design element editing, model viewing, model comparison, etc. in the design process, and users of different design types correspond to different design contents, such as users responsible for different professional models. Therefore, the collected interaction data has high-dimensional behavioral characteristics. If a simple behavioral model analysis is performed on users based on traditional methods, it will lead to large individual differences and it will be difficult to classify users by behavioral characteristics. Based on this, the present invention performs data dimensionality reduction analysis on user interaction data through the PCA principal component analysis method, and further screens out the principal component characteristics for subsequent user clustering and grouping.

[0048] According to an embodiment of the present invention, the principal component features of each user are clustered based on the K-means algorithm to form multiple designed user groups, specifically: Construct a clustering model based on K-means and set K center points to initialize the clustering model; Import the principal component features of each user into the clustering model, use the principal component features as clustering sample data, use each user as a clustering unit, calculate the distance from the principal component features to the center point, and group the sample data based on the distance; Repeatedly calculate the new center point and compare the new center point with the original center point until the center point no longer moves. At this time, record the current clustering result; According to the current clustering results, users are clustered and grouped to form multiple designed user groups.

[0049] It should be noted that the clustering result includes the grouping of the principal component features, and corresponds to the grouping of the users. The user groups are designed to be grouped based on the user interaction behavior features, and the interaction features of the users in each group have certain similarities.

[0050] According to an embodiment of the present invention, within a preset historical time period, network interaction data and server preloading information of a design user group are collected, and real-time resource loading capability evaluation of the design user group is performed through the network interaction data and the server preloading information to generate resource loading capability information, specifically: In a preset historical time period, a design user group is used as an analysis unit to collect network interaction data and server preloading information of a design user group; The network interaction data includes port information, network status, and data transmission rate information; The server preloading information includes historical loading data volume, loading waiting time, and preloading success rate information; Through network interaction data and server preloading information, a design user group is evaluated based on network latency, network rate, and average loaded data volume, and resource loading capacity information is generated.

[0051] It should be noted that the preloading success rate information, i.e., the ratio of the amount of successfully preloaded data to the total load, reflects the preloading data capacity within a preset historical time period. Based on different user terminals and different network conditions, when users open, edit, and design building models, the preloading conditions of the models are different. The resource loading capacity information is used to reflect the average resource loading capacity of a design user group. Through this capacity information, targeted resource loading optimization solution analysis can be performed in the subsequent period.

[0052] According to an embodiment of the present invention, the target building model is divided into multiple design areas, and real-time design interaction instructions of a design user group are collected within a preset historical time period. The design areas browsed, designed, and edited by users are marked and counted through the interaction instructions, and key areas and non-key areas are screened out. Combined with the resource loading capability information, the preloading scheme of the design user group is optimized to generate a model loading scheme based on intelligent optimization, specifically: Through the architectural design platform, the target building model is divided into multiple design areas; In a preset historical time period, the real-time design interaction instructions of all users in a design user group are collected, and the interaction instructions are identified. Through instruction identification, the design areas corresponding to the user's browsing, design, and editing behaviors are marked and counted, and the interaction frequency of different design areas is analyzed to mark high-frequency interaction areas and low-frequency interaction areas; The high-frequency interaction areas and low-frequency interaction areas are marked as key areas and non-key areas respectively.

[0053] It should be noted that the design and editing focuses of different design users are in different design areas.

[0054] According to an embodiment of the present invention, the method of optimizing the preloading scheme for the design user group in combination with the resource loading capability information and generating a model loading scheme based on intelligent optimization further includes: In a preset historical time period, real-time design interaction instructions of a design user group are collected, instruction recognition and user behavior analysis are performed on the interaction instructions, the instructions are associated with the user behavior, the instructions are sorted based on the time dimension of the collected interaction instructions, and a user behavior sequence is generated based on the instruction sorting result; In a preset historical time period, based on whether the design interface in which the user is located is a key area or a non-key area during the interactive design, the user behavior sequence is divided to generate a first behavior sequence and a second behavior sequence; Build a prediction model based on LSTM and import historical user behavior sequence data into the prediction model for model training; A prediction cycle is set, and the first behavior sequence and the second behavior sequence are respectively imported into the prediction model to predict user behavior, and a first predicted behavior and a second predicted behavior are obtained; According to the first predicted behavior and the second predicted behavior, the design behavior pattern of the design user group is analyzed, and the loading plan analysis is performed on the key areas and non-key areas. Combined with the resource loading capability information, the loading model file, loading file type, and file loading amount parameters are set, the original loading plan is optimized, and a model loading plan for the design user group is generated.

[0055] It should be noted that the interactive instructions include client control instructions when the user browses, designs, and edits the corresponding design area. Through instruction recognition, pattern recognition of user behavior can be performed. In the present invention, by identifying and serializing instructions, a corresponding user behavior sequence can be generated, and user behavior prediction is performed based on a prediction model. The user behavior sequence corresponds one-to-one to the instruction sequence. The first behavior sequence corresponds to the behavior data generated when the user performs interactive design in the key area, and the second behavior sequence corresponds to the behavior data generated when the user performs interactive design in the non-key area. The first predicted behavior and the second predicted behavior are design prediction behaviors for the design user group in the key area and non-key area of ​​the model, and the user's subsequent preloading solution analysis.

[0056] It is worth mentioning that in the actual architectural design process, different users have different frequencies of editing, using, and updating content for different design areas, and different design focuses. For example, in the target architectural model, there are multiple professional models, and different users or users of different user groups have different key design areas for the model. There are also certain differences in the design areas of each client loading model, design model, and editing model. Therefore, in view of the differences between different users, the present invention analyzes the interactive characteristics and clusters the users, generates corresponding key and non-key design areas for each user group, and generates sequence data based on user design behavior based on the correlation between interactive instructions and behavior models. According to the sequence data, the LSTM prediction model in the present invention can be combined to evaluate user behavior, and the pre-loaded data can be adjusted to improve the comprehensive design efficiency of large-scale and different professional design users, improve the intelligence and automation of the platform, shorten the design cycle, and improve the efficiency of collaborative design. At the same time, the intelligent optimization of model loading is effectively carried out, so that each design user group has a different loading scheme to improve the overall design efficiency.

[0057] The key areas and non-key areas are used to determine whether the user group performs design interactions in the key high-frequency design model. The key areas are screened based on the user's interaction data over a period of time, specifically to the marked areas within the architectural design.

[0058] It is worth mentioning that traditional preloading solutions often simply analyze user design preferences and load commonly used tool models. However, for large-scale collaborative design and design platforms with multi-professional architectural models, traditional methods are inefficient and difficult to meet modern design needs.

[0059] The third aspect of the present invention also provides a computer-readable storage medium, which includes an architectural design program based on intelligent optimization. When the architectural design program based on intelligent optimization is executed by a processor, the steps of the architectural design method based on intelligent optimization as described in any one of the above items are implemented.

[0060] The present invention discloses an architectural design system based on intelligent optimization. The present invention collects user design interaction data through an architectural design platform, and counts the design area, component, time and frequency to form user design interaction characteristics. The PCA principal component analysis method is used to extract the principal components of the user interaction characteristics, and the K-means algorithm is used for clustering to form multiple design user groups. Within a preset time period, the real-time resource loading capability of the design user group is evaluated, and the preloading scheme is optimized in combination with the resource loading capability information to generate a model loading scheme based on intelligent optimization. Through the present invention, it is possible to improve the comprehensive design efficiency of large-scale and different professional design users, enhance the intelligence and automation of the platform, shorten the design cycle, and improve the efficiency of collaborative design.

[0061] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as: multiple units or components can be combined, or can be integrated into another system, or some features can be ignored or not executed. In addition, the coupling, direct coupling, or communication connection between the components shown or discussed can be through some interfaces, and the indirect coupling or communication connection of the devices or units can be electrical, mechanical or other forms.

[0062] The units described above as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units; they may be located in one place or distributed on multiple network units; some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0063] In addition, all functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may be separately used as a unit, or two or more units may be integrated into one unit; the above-mentioned integrated units may be implemented in the form of hardware or in the form of hardware plus software functional units.

[0064] A person of ordinary skill in the art can understand that: all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: a mobile storage device, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

[0065] Alternatively, if the above-mentioned integrated unit of the present invention is implemented in the form of a software function module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the embodiment of the present invention can be essentially or partly reflected in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as mobile storage devices, ROM, RAM, magnetic disks or optical disks.

[0066] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A building design method based on intelligent optimization, characterized in that: include: Through the architectural design platform, user design interaction data is obtained. Through the design interaction data, the design areas, design components, design time, and design frequency of different users are counted to obtain user design interaction characteristics; Based on the user design interaction features, the principal component selection of the user interaction features is performed through the PCA principal component analysis method to generate the principal component features of each user; Based on the K-means algorithm, the principal component features of each user are clustered to form multiple designed user groups; In a preset historical time period, network interaction data and server preloading information of a design user group are collected, and real-time resource loading capability evaluation of the design user group is performed based on the network interaction data and server preloading information to generate resource loading capability information; The target building model is divided into multiple design areas. Within the preset historical time period, the real-time design interaction instructions of a design user group are collected. The design areas browsed, designed, and edited by users are marked and counted through the interaction instructions, and the key areas and non-key areas are screened out. Combined with the resource loading capacity information, the pre-loading plan for the design user group is optimized to generate a model loading plan based on intelligent optimization.

2. The architectural design method based on intelligent optimization according to claim 1, characterized in that: The user design interaction data is obtained through the architectural design platform, and the design areas, design components, design time, and design frequency of different users are counted through the design interaction data to obtain user design interaction features, which are specifically: Build a cloud computing-based architectural design platform; Data transmission between multiple user terminals through the building design platform; Based on the network transmission between the architectural design platform and multiple user terminals, the user design interaction data is obtained; According to the design areas, design components, design time and design frequency of different users, data statistics and initial user characteristics analysis are performed on the user design interaction data to obtain the user design interaction characteristics.

3. The architectural design method based on intelligent optimization according to claim 2 is characterized in that: The user-designed interaction features are used to select the principal components of the user interaction features through the PCA principal component analysis method to generate the principal component features of each user, specifically: Standardize the data of user design interaction features; Based on the PCA analysis method, the characteristic variables of user design interaction features are extracted and the covariance matrix is ​​constructed; Perform eigenvalue decomposition on the covariance matrix to form eigenvalues ​​and corresponding eigenvectors; Sorting is performed based on eigenvalues, and the first N eigenvectors are selected as principal components according to preset standards; A new feature space is constructed based on the selected principal component feature vectors. Through the new feature space, the user interaction features are reconstructed and principal component features are generated.

4. The architectural design method based on intelligent optimization according to claim 3 is characterized in that: Based on the K-means algorithm, the principal component features of each user are clustered to form multiple designed user groups, specifically: Construct a clustering model based on K-means and set K center points to initialize the clustering model; Import the principal component features of each user into the clustering model, use the principal component features as clustering sample data, use each user as a clustering unit, calculate the distance from the principal component features to the center point, and group the sample data based on the distance; Repeatedly calculate the new center point and compare the new center point with the original center point until the center point no longer moves. At this time, record the current clustering result; According to the current clustering results, users are clustered and grouped to form multiple designed user groups.

5. The architectural design method based on intelligent optimization according to claim 4 is characterized in that: In the preset historical time period, network interaction data and server preloading information of a design user group are collected, and real-time resource loading capability evaluation of the design user group is performed through the network interaction data and server preloading information to generate resource loading capability information, specifically: In a preset historical time period, a design user group is used as an analysis unit to collect network interaction data and server preloading information of a design user group; The network interaction data includes port information, network status, and data transmission rate information; The server preloading information includes historical loading data volume, loading waiting time, and preloading success rate information; Through network interaction data and server preloading information, a design user group is evaluated based on network latency, network rate, and average loaded data volume, and resource loading capacity information is generated.

6. The architectural design method based on intelligent optimization according to claim 5 is characterized in that: The target building model is divided into multiple design areas, and real-time design interaction instructions of a design user group are collected within a preset historical time period. The design areas browsed, designed, and edited by users are marked and counted through the interaction instructions, and key areas and non-key areas are screened out. Combined with the resource loading capacity information, the pre-loading scheme of the design user group is optimized to generate a model loading scheme based on intelligent optimization, specifically: Through the architectural design platform, the target building model is divided into multiple design areas; In a preset historical time period, the real-time design interaction instructions of all users in a design user group are collected, and the interaction instructions are identified. Through instruction identification, the design areas corresponding to the user's browsing, design, and editing behaviors are marked and counted, and the interaction frequency of different design areas is analyzed to mark high-frequency interaction areas and low-frequency interaction areas; The high-frequency interaction areas and low-frequency interaction areas are marked as key areas and non-key areas respectively.

7. The architectural design method based on intelligent optimization according to claim 6 is characterized in that: The method combines the resource loading capability information to optimize the preloading solution for the design user group and generates a model loading solution based on intelligent optimization, and further includes: In a preset historical time period, real-time design interaction instructions of a design user group are collected, instruction recognition and user behavior analysis are performed on the interaction instructions, the instructions are associated with the user behavior, the instructions are sorted based on the time dimension of the collected interaction instructions, and a user behavior sequence is generated based on the instruction sorting result; In a preset historical time period, based on whether the design interface in which the user is located is a key area or a non-key area during the interactive design, the user behavior sequence is divided to generate a first behavior sequence and a second behavior sequence; Build a prediction model based on LSTM and import historical user behavior sequence data into the prediction model for model training; A prediction cycle is set, and the first behavior sequence and the second behavior sequence are respectively imported into the prediction model to predict user behavior, and a first predicted behavior and a second predicted behavior are obtained; According to the first predicted behavior and the second predicted behavior, the design behavior pattern of the design user group is analyzed, and the loading plan analysis is performed on the key areas and non-key areas. Combined with the resource loading capability information, the loading model file, loading file type, and file loading amount parameters are set, the original loading plan is optimized, and a model loading plan for the design user group is generated.

8. An architectural design system based on intelligent optimization, characterized in that: The system includes: a memory and a processor, wherein the memory includes an architectural design program based on intelligent optimization, and when the architectural design program based on intelligent optimization is executed by the processor, the following steps are implemented: Through the architectural design platform, user design interaction data is obtained. Through the design interaction data, the design areas, design components, design time, and design frequency of different users are counted to obtain user design interaction characteristics; Based on the user design interaction features, the principal component selection of the user interaction features is performed through the PCA principal component analysis method to generate the principal component features of each user; Based on the K-means algorithm, the principal component features of each user are clustered to form multiple designed user groups; In a preset historical time period, network interaction data and server preloading information of a design user group are collected, and real-time resource loading capability evaluation of the design user group is performed based on the network interaction data and server preloading information to generate resource loading capability information; The target building model is divided into multiple design areas. Within the preset historical time period, the real-time design interaction instructions of a design user group are collected. The design areas browsed, designed, and edited by users are marked and counted through the interaction instructions, and the key areas and non-key areas are screened out. Combined with the resource loading capacity information, the pre-loading plan for the design user group is optimized to generate a model loading plan based on intelligent optimization.

9. The architectural design system based on intelligent optimization according to claim 8, characterized in that: The user design interaction data is obtained through the architectural design platform, and the design areas, design components, design time, and design frequency of different users are counted through the design interaction data to obtain user design interaction features, which are specifically: Build a cloud computing-based architectural design platform; Data transmission between multiple user terminals through the building design platform; Based on the network transmission between the architectural design platform and multiple user terminals, the user design interaction data is obtained; According to the design areas, design components, design time and design frequency of different users, data statistics and initial user characteristics analysis are performed on the user design interaction data to obtain the user design interaction characteristics.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes an architectural design program based on intelligent optimization. When the architectural design program based on intelligent optimization is executed by a processor, the steps of the architectural design method based on intelligent optimization as described in any one of claims 1 to 7 are implemented.

Images