Building design project intelligent evaluation method and system based on multi-dimensional indexes

Abstract

The invention relates to the field of building design, and discloses a building design project intelligent evaluation method and system based on multi-dimensional indexes, and the method comprises the following steps: S1, collecting building design data from multiple sources; s2, preprocessing the collected various data; s3, constructing a multi-dimensional evaluation system, and generating a comprehensive evaluation result; s4, adjusting the weight of each evaluation dimension; s5, generating a comprehensive evaluation report; the system comprises a data acquisition module, a data preprocessing module, an evaluation system construction module, a reinforcement learning adjustment module, a real-time feedback mechanism module, a comprehensive evaluation report generation module and a data fusion module. According to the method, data fusion is carried out through a multi-dimensional index evaluation system and in combination with a deep learning algorithm, comprehensive and accurate evaluation of the building design project is achieved, and compared with a technical scheme of general single-dimensional evaluation in the prior art, the problem that part of important factors are ignored during evaluation in the traditional technology is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of architectural design, and specifically to an intelligent evaluation method and system for architectural design projects based on multi-dimensional indicators. Background Art

[0002] At present, the evaluation of architectural design projects mainly relies on a variety of technical means and methods. Common evaluation systems include functional evaluation, economic evaluation, and environmental impact evaluation, etc. Architectural design evaluation usually collects information from multiple data sources such as BIM, environmental monitoring, and user behavior data to form an all-round evaluation of the construction project. These evaluation methods are based on static data in the design stage and combine some environmental impact evaluation, energy efficiency analysis, etc., aiming to provide a preliminary judgment on the quality, performance, and later operation of the architectural design. In traditional architectural design evaluation systems, the system helps decision-makers identify the advantages and potential improvement spaces in the design by integrating and analyzing various indicators. Nevertheless, these evaluation systems still play an important role in practical applications, especially in the preliminary design and project planning stages, providing useful reference information.

[0003] However, the architectural design evaluation methods of the prior art still have certain limitations. First of all, the evaluation system basically relies on fixed evaluation dimensions and weight settings, and this method fails to flexibly respond to the changes that occur during the actual operation of the project. For example, as the building project is used, the importance of certain dimensions may change, and the prior art usually cannot dynamically adjust the weights of the evaluation dimensions according to real-time data. In addition, the existing evaluation methods usually focus on the static data in the design stage and lack continuous monitoring and feedback on the real-time data generated during the actual operation of the building project. This makes the evaluation results unable to reflect the performance of the architectural design in the actual use process in real time, thus affecting the later design optimization and decision support. Finally, due to the low degree of automation, the existing evaluation systems still rely on manual intervention, and the optimization efficiency and decision support capabilities are relatively limited. Therefore, the prior art has obvious deficiencies in dynamic adaptability, real-time performance, and automation. Summary of the Invention

[0004] Aiming at the deficiencies of the prior art, the present invention provides an intelligent evaluation method and system for architectural design projects based on multi-dimensional indicators, which solves the problems of the existing architectural design evaluation system in the aspects of evaluation dimension, weight dynamic adjustment, real-time data feedback, and automation optimization.

[0005] To achieve the above objectives, the present invention is realized through the following technical solutions: An intelligent evaluation method for architectural design projects based on multi-dimensional indicators, including the following steps: S1. Collect architectural design data from multiple sources, including BIM data, environmental impact data, user behavior data, project cost data, and building performance data; S2. Preprocess various types of collected data, including data cleaning, standardization, and mapping, and integrate data from different sources into a unified format; S3. Construct a multi-dimensional evaluation system, set evaluation dimensions and quantitative indicators, and deeply fuse multi-dimensional data through deep learning algorithms to generate a comprehensive evaluation result; S4. Dynamically adjust the weights of each evaluation dimension through reinforcement learning algorithms, optimize the weight allocation based on historical design data and real-time feedback, and collect the operation data of construction projects in real time to automatically adjust the evaluation weights; S5. Generate a comprehensive evaluation report and provide design optimization suggestions to assist design decisions.

[0006] Preferably, in step S1, the BIM data includes construction parameters such as the spatial layout, structural form, material selection, and construction technology of the building design; The environmental impact data includes building energy efficiency, carbon emissions, climate impact, and resource consumption data, and real-time data is collected through Internet of Things devices and a building management system.

[0007] Preferably, in step S2, data cleaning includes removing outliers, filling in missing data, and removing redundant information; The standardization process is carried out by converting data from different sources into a unified dimension, ensuring that various types of data can be compared on the same scale, and using the Z-score standardization method to normalize the data.

[0008] Preferably, in step S3, the evaluation dimensions include design functionality, sustainability, economy, user experience, and environmental impact, and each evaluation dimension is scored by setting corresponding quantitative indicators; The deep learning algorithm uses convolutional neural networks and recurrent neural networks to extract features from different types of data, generate a comprehensive evaluation result, and perform weighted fusion on the scores of each evaluation dimension.

[0009] Preferably, the process of weighted fusion can be expressed by the following formula: ; where, is the comprehensive evaluation result; is the weight of the th data source; is the feature extracted from the th data source.

[0010] Preferably, in step S4, the reinforcement learning algorithm adopts the Q-learning algorithm, and dynamically adjusts the weight allocation of each evaluation dimension by calculating the current weight and corresponding score of each evaluation dimension; The real-time feedback mechanism includes collecting real-time energy efficiency data and user behavior data of a construction project, connecting through Internet of Things devices to a building management system, and adjusting weights in real time and optimizing the evaluation results.

[0011] Preferably, the specific formula of the Q-learning algorithm is as follows: ; Wherein, is the current state and action of value; is the learning rate, which controls the learning speed of the model; is the current state and weight corresponding reward value; is the discount factor, indicating the importance of future rewards; in, the maximum value of all possible actions.

[0012] Preferably, in step S5, the deep learning algorithm includes: A convolutional neural network, which is used to process building space layout data; A recurrent neural network, which is used to process time series data; Weightedly fuse the features extracted by both to generate a comprehensive evaluation result.

[0013] Preferably, the weighted fusion for generating the comprehensive evaluation result is represented by the following formula: ; Wherein, is the final comprehensive evaluation result; is the weight of the th data source; is the evaluation result of the th data source.

[0014] An intelligent evaluation system for building design projects based on multi-dimensional indicators includes: A data collection module, which is used to collect building design data from multiple sources, including BIM data, environmental impact data, user behavior data, project cost data, and building performance data; A data preprocessing module, which is connected to the data collection module and is used to clean and standardize the collected building design data; The evaluation system construction module, which is connected to the data preprocessing module, is used to construct a multi-dimensional evaluation system, set evaluation dimensions and quantitative indicators, and deeply fuse the evaluation results of different data sources through deep learning algorithms to generate a comprehensive evaluation result; The reinforcement learning adjustment module, which is connected to the evaluation system construction module, is used to dynamically adjust the weights of the evaluation dimensions, calculate the current weights of each evaluation dimension through reinforcement learning algorithms, and automatically optimize the weight allocation according to real-time feedback; The real-time feedback mechanism module, which is connected to the data acquisition module, is used to collect the operation data of the construction project in real time, including energy efficiency data and user behavior data, and feedback it to the reinforcement learning adjustment module through Internet of Things devices to optimize the weights of the evaluation dimensions; The comprehensive evaluation report generation module, which is connected to the evaluation system construction module, is used to generate a comprehensive evaluation report based on the evaluation results processed by deep learning algorithms and provide design optimization suggestions; The data fusion module, which is connected to the evaluation system construction module, is used to perform weighted fusion on the evaluation results from different data sources and generate the final comprehensive evaluation score.

[0015] The present invention provides an intelligent evaluation method and system for building design projects based on multi-dimensional indicators. It has the following beneficial effects: 1. Through the evaluation system of multi-dimensional indicators and combined with deep learning algorithms for data fusion, the present invention achieves a comprehensive and accurate evaluation of building design projects. Compared with the technical solutions that usually use single-dimensional evaluation in the prior art, this solution can consider multiple dimensions such as design functionality, sustainability, economy, user experience, and environmental impact at the same time, providing a more comprehensive analysis perspective and solving the problem that some important factors are ignored in the evaluation of traditional technologies.

[0016] 2. By introducing reinforcement learning algorithms to dynamically adjust the weights of the evaluation dimensions, the present invention achieves the adaptive ability of the evaluation system. Compared with the evaluation methods with fixed weights in the prior art, this solution can automatically optimize the weight allocation according to the operation data and historical feedback of the construction project in real time, thereby improving the accuracy and reliability of the evaluation results and solving the limitation that fixed weights cannot adapt to project changes.

[0017] 3. By fusing multi-source information such as BIM data, real-time environmental monitoring data, and user behavior data, the present invention achieves the generation of more accurate design optimization suggestions. Different from the evaluation that only relies on static design data in traditional technologies, this solution enables the evaluation results to dynamically reflect the performance of building design in actual operation through real-time collected data, solving the dilemma of the mismatch between static evaluation and actual operation.

[0018] 4. By combining deep learning algorithms with reinforcement learning algorithms, the present invention automatically extracts optimization information from historical data and real-time data, achieving a highly automated generation of design optimization reports. Different from the way of manually adjusting optimization strategies in the prior art, this solution automatically generates comprehensive evaluation reports and optimization suggestions through algorithms, greatly reducing manual intervention, improving work efficiency, and avoiding errors caused by human biases at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 is a flowchart of the method steps of the present invention; Figure 2 is a system module architecture diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings of the present invention specification. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0021] Please refer to the attached Figure 1 , the embodiments of the present invention provide an intelligent evaluation method for building design projects based on multi-dimensional indicators, including the following steps: S1. Collect building design data from multiple sources, including BIM data, environmental impact data, user behavior data, project cost data, and building performance data.

[0022] Specifically, step S1 includes collecting building design data from multiple sources to ensure that the system can comprehensively evaluate all aspects of building design. Specifically, the data sources include building information model (BIM) data, environmental impact data, user behavior data, project cost data, and building performance data, etc. These data cover multiple aspects such as the spatial layout, material selection, structural form, construction technology, energy efficiency, carbon emissions, and resource consumption of building design projects, reflecting the comprehensive performance and impact of building projects. The data collection not only comes from static design parameters but also includes dynamic operation data and real-time feedback information to ensure the evaluation ability of the system during the building life cycle.

[0023] Generally, as an important information carrier for building design, BIM data mainly contains design parameters such as the spatial layout, structural form, material selection, and construction technology of building projects. BIM data can be gradually accumulated from the building design stage through building information modeling technology and cover multiple stages such as design, construction, and operation. Through BIM data, the three-dimensional model of the building can be intuitively displayed, and then detailed spatial layout and construction parameters can be provided for subsequent intelligent evaluation.

[0024] As an option, the environmental impact data includes information such as the energy efficiency, carbon emissions, climate impact, and resource consumption of the construction project. These data can reflect the impact of the building design on the environment and help evaluate the sustainability of the construction project in terms of energy consumption and environmental protection. In this embodiment, the environmental impact data can be collected in real time through a building management system (BMS) connected to Internet of Things (IoT) devices. The IoT devices can monitor various environmental parameters inside and outside the building in real time, including energy efficiency, temperature, humidity, lighting intensity, etc., and transmit them to the intelligent evaluation system through the BMS for real-time evaluation of the actual operation effect of the building design.

[0025] In a possible implementation, the user behavior data includes the actual usage data of the construction project, such as the behavior patterns of users, the frequency of space usage, the equipment usage conditions, etc. These data can be collected through sensors or user devices (such as smartphones, smart bracelets, etc.) and integrated with other data sources. The user behavior data is of great significance for evaluating the humanization, comfort, and functionality of the building design.

[0026] In addition, the project cost data, as an important basis for evaluating the economy of the building design, includes construction costs, operation and maintenance costs, energy-saving benefits, etc. These data can provide support for the financial and economic evaluation of the construction project. By analyzing the project cost data, it is possible to judge the feasibility and benefits of the design scheme within the budget.

[0027] Specifically, the building performance data includes performance indicators such as the actual energy efficiency, environmental adaptability, and seismic resistance of the building. These data can be collected and evaluated through the operation data of the building, providing a basis for subsequent design optimization and improvement.

[0028] Through the data collection from the above multiple sources, the present invention can ensure that comprehensive design parameters and actual operation conditions are considered when evaluating the building design, and thus provide a scientific and accurate basis for design optimization and decision-making.

[0029] In some embodiments, the data collection module uses automated data collection devices and is seamlessly connected to the building management system or other information systems. This method can ensure the real-time and accuracy of the data collection process, reduce manual intervention, and improve the efficiency and quality of data processing.

[0030] S2. Preprocess the collected various types of data, including data cleaning, standardization, and mapping, and integrate the data from different sources into a unified format.

[0031] Specifically, in step S2, data cleaning first includes removing outliers, which may be caused by errors in the data collection process, data entry errors, or interference during transmission. The purpose of removing outliers is to ensure the quality of the data and reduce the impact of inaccurate data on subsequent evaluations. The detection of outliers can be carried out through statistical methods, such as methods based on standard deviation or box plots, to ensure the rationality and validity of the data.

[0032] As an option, data cleaning also includes filling in missing data. Since there may be some missing data during the building design process, the filling of missing data can be completed through interpolation methods, regression analysis, or prediction methods based on similar data. Common filling methods include mean filling, interpolation methods (such as linear interpolation), and filling methods based on machine learning. According to the characteristics and missing situations of the data, appropriate filling methods can be selected to ensure the integrity of the data during subsequent analysis.

[0033] In addition, data cleaning also includes removing redundant information, which may come from duplicate records, useless variables, or irrelevant features. Removing redundant information helps reduce the burden of data processing, improve computational efficiency, and avoid unnecessary interference with the analysis results. The removal of redundant information can be achieved through feature selection methods, correlation analysis, or principal component analysis (PCA) technical means. The goal of PCA is to project the data from a high-dimensional space to a low-dimensional space through dimensionality reduction and reduce the redundancy of the data: ; where is the projection matrix, is the original data, is the data after dimensionality reduction.

[0034] After data cleaning, the data from different sources are then standardized. The purpose of standardization is to convert the data from different sources into a unified dimension so that different types of data can be compared and integrated on the same scale. Commonly used standardization methods include Z-score standardization and Min-Max normalization. Specifically, the Z-score standardization method compares the deviation of each data point from its mean and divides it by the standard deviation. The formula is as follows: ; where is the value after standardization; is the original data; is the mean of the data; is the standard deviation of the data.

[0035] Through the above standardization process, all data from different sources will be evaluated under the same standard, eliminating the inconsistency caused by different dimensions between data. In some embodiments, the selection of the standardization method can be adjusted according to the characteristics of the specific data. For example, if the data range is small, Min-Max normalization may be more appropriate; while for data with a wider distribution, Z-score normalization can more effectively reduce the impact of extreme values.

[0036] In a possible implementation, data preprocessing also includes data mapping. The purpose of data mapping is to map data from different sources to a unified feature space so that subsequent deep learning algorithms can effectively extract features and perform weighted fusion. Data mapping usually relies on feature engineering techniques. By transforming and mapping the data, it is ensured that the input data can retain the original information to the greatest extent and meet the requirements of subsequent models.

[0037] Specifically, dimensionality reduction techniques such as PCA (Principal Component Analysis) may be used during the mapping process, or domain knowledge may be utilized to map the data so that data from different sources can be compared on the same dimension. Through these methods, it can be ensured that the data has higher usability and expressive ability in subsequent steps.

[0038] S3. Construct a multi-dimensional evaluation system, set evaluation dimensions and quantitative indicators, and perform deep fusion on multi-dimensional data through deep learning algorithms to generate a comprehensive evaluation result.

[0039] Specifically, in step S3, the construction of the evaluation system is based on multiple dimensions, including dimensions such as design functionality, sustainability, economy, user experience, and environmental impact. Each evaluation dimension is scored through a set of quantitative indicators, so as to quantitatively reflect the performance of different aspects of the building design project. These dimensions are the basis for the system to comprehensively and accurately evaluate the building design, ensuring the system's multi-angle analysis ability and comprehensive evaluation ability.

[0040] Specifically, the design functionality in the evaluation dimension mainly focuses on whether the building design meets its predetermined functions and usage requirements. Sustainability evaluates the performance of the design scheme in terms of environmental protection, energy efficiency, and resource utilization. The economic dimension measures the cost-effectiveness of the design scheme, including the total construction cost and long-term operation and maintenance costs of the project. User experience evaluates the impact of the building space design on the comfort, convenience, and psychological perception of users. Environmental impact involves the impact of the building design on the ecological environment, including parameters such as carbon emissions and energy consumption.

[0041] As an option, deep learning algorithms play a key role in this step. Through deep learning algorithms such as convolutional neural networks (CNNs) and recurrent neural networks (RNNs), feature extraction can be performed on building space layout data and time series data respectively.

[0042] A convolutional neural network (CNN) can extract features from spatial data in the following way: ; where represents the feature after convolution, is the convolutional kernel, represents the convolution operation, is the input data, is the bias term, is the activation function.

[0043] For time series data, the RNN model processes time series data through a recursive relationship: ; where is the hidden state at time , is the input data, and are weight matrices, is the bias term, is the activation function.

[0044] After feature extraction is performed on spatial data and time series data by applying CNN and RNN respectively, the system performs weighted fusion, and this process is achieved through the following formula: ; where is the final comprehensive evaluation result, is the weight of the th evaluation dimension, is the th dimension feature Specifically, when processing spatial layout data, CNN can automatically extract the spatial features of a building, such as the functional zoning of the building, traffic paths, etc.; while RNN is mainly used to process time series data, especially dynamic data related to building operation, such as energy efficiency, temperature changes, and other environmental parameters. Through the combination of these two neural networks, the system can process spatial data and time series data simultaneously and generate their respective feature vectors.

[0045] In some embodiments, when a convolutional neural network (CNN) processes spatial layout data, it extracts hierarchical features of the building space layer by layer through multiple convolutional layers. These features can reflect structural relationships in architectural design, the rationality of spatial layout, and other aspects. At the same time, a recurrent neural network (RNN) is used to process time series data related to the operation of a building project, such as equipment usage, building energy efficiency, etc. The time series processing ability of the RNN enables the system to capture the performance and change trends of architectural design at different time stages.

[0046] Specifically, the process of generating a comprehensive evaluation result is achieved by weighted fusion of data from different dimensions. Each evaluation dimension is scored through quantitative indicators, and the scores of all dimensions are weighted and fused according to their importance. The formula for weighted fusion is as follows: ; where is the comprehensive evaluation result; is the weight of the th data source; is the th feature extracted from the data source.

[0047] In some embodiments, the weight can be automatically learned through a deep learning algorithm. The system dynamically adjusts the importance of evaluation dimensions based on historical data and real-time feedback to ensure that the comprehensive evaluation result can accurately reflect the actual performance of the project. The weighted fusion process comprehensively considers the results of all evaluation dimensions and generates a comprehensive and integrated evaluation result.

[0048] Through this weighted fusion method, the system can comprehensively evaluate a building design project from multiple perspectives, rather than being limited to a single dimension. The evaluation result will provide a scientific basis for building design decisions, helping designers optimize the design plan and improve the overall performance of the building project.

[0049] S4. Dynamically adjust the weights of each evaluation dimension through a reinforcement learning algorithm, optimize the weight allocation based on historical design data and real-time feedback, and collect the operation data of the building project in real time to automatically adjust the evaluation weights.

[0050] Specifically, in step S4, it generates a comprehensive evaluation result through a deep learning algorithm and integrates the scores of different evaluation dimensions through weighted fusion. However, due to the complexity of building design projects and the variability of the environment, a static evaluation system is difficult to meet the dynamic requirements of the system. Therefore, in step S4, a reinforcement learning algorithm is used to dynamically adjust the weights of evaluation dimensions to ensure that the system can adapt to new data and environmental changes at different design stages or operation stages.

[0051] In this embodiment, the reinforcement learning algorithm adopts the Q-learning algorithm to achieve self-optimization during the process of weight allocation in the evaluation dimension. The Q-learning algorithm dynamically adjusts the weights of each evaluation dimension by calculating the current weights and corresponding scores of each evaluation dimension, thereby optimizing the comprehensive evaluation result. The core of the Q-learning algorithm is to guide the agent (i.e., the evaluation system) to select the optimal evaluation dimension weights through the reward function, so that the evaluation result can better conform to the actual situation.

[0052] Specifically, the update formula of the Q-learning algorithm is as follows: ; where is the current state and action of value; is the learning rate, which controls the learning speed of the model; is the current state and weight corresponding reward value; is the discount factor, indicating the importance of future rewards; in, the maximum value of all possible actions.

[0053] The core of this formula is that, based on the current state and the selected weights, the system calculates the reward value , and adjusts the weights of the evaluation dimensions by updating the Q value. By continuously iterating this process, the system can gradually optimize the weight allocation of each evaluation dimension after multiple interactions, so that the final evaluation result better meets the actual requirements.

[0054] As an option, the real-time feedback mechanism plays a crucial role in this step. The energy efficiency data, user behavior data, and other real-time parameters of the construction project are collected in real time, connected through Internet of Things devices and the building management system (BMS), and the weights of the evaluation dimensions are adjusted in real time. This feedback mechanism can ensure that the system dynamically adjusts the evaluation weights according to the actual situation during the operation stage of the construction project, thereby optimizing the evaluation result. This real-time update ability not only improves the accuracy of the evaluation result but also enhances the adaptability of the system.

[0055] In some embodiments, the feedback mechanism may include monitoring parameters such as the energy efficiency, temperature changes, and air quality of a construction project, and adjusting the weights based on this real-time data. If the system detects poor energy efficiency of the building or problems in certain user experience aspects, the feedback mechanism will prompt the Q-learning algorithm to adjust the weights of relevant evaluation dimensions, thereby paying more attention to the direction of energy efficiency optimization or user experience improvement. In this way, the evaluation results can reflect the actual operating conditions of the building in a timely manner, avoiding errors caused by static evaluations in the design stage.

[0056] S5. Generate a comprehensive evaluation report and provide design optimization suggestions to assist in design decision-making.

[0057] Specifically, in step S5, it involves generating a comprehensive evaluation report and providing design optimization suggestions as part of the system output. This step generates a comprehensive and accurate evaluation report through the multi-dimensional evaluation results obtained in the previous steps, combined with the optimization of deep learning algorithms and reinforcement learning algorithms. The weighted fusion of the comprehensive evaluation results is represented by the following formula: ; where, is the final comprehensive evaluation result; is the weight of the th data source; is the evaluation result of the th data source.

[0058] This report not only summarizes the evaluation scores of various aspects of the building design project but also provides specific optimization suggestions for possible deficiencies. Through these suggestions, designers can further improve the design scheme and enhance the overall performance and user experience of the construction project.

[0059] In the aforementioned step S4, the weights of each evaluation dimension were dynamically adjusted through the reinforcement learning algorithm, optimizing the evaluation results. Therefore, in step S5, the system can generate the final comprehensive evaluation report based on the optimized evaluation results. The report includes the detailed scores of each evaluation dimension and the weighted comprehensive score, which provide a quantitative basis for design optimization. The comprehensive evaluation report will provide detailed references for all relevant parties of the building design project to assist in the decision-making process.

[0060] Specifically, the generation process of the comprehensive evaluation report is based on the comprehensive evaluation results obtained by the system in step S4. First, the system integrates the scores of each evaluation dimension according to the weighted fusion method to generate a comprehensive comprehensive score. This score reflects the performance of the building design project in different dimensions, such as design functionality, sustainability, economy, user experience, and environmental impact.

[0061] In some embodiments, the comprehensive evaluation report not only includes the comprehensive score, but may also include a detailed analysis of each evaluation dimension, pointing out the reasons for the lower scores in each dimension. For example, if the score is low in the sustainability dimension, the report may indicate that the design solution has deficiencies in energy efficiency or resource utilization, and propose specific improvement measures, such as optimizing the energy efficiency of buildings and reducing carbon emissions. Such a report structure enables designers to quickly identify the shortcomings in the design and take corresponding optimization measures.

[0062] Specifically, the generation of the comprehensive evaluation report can be achieved through the following process: The system first calculates the comprehensive score of the building design project based on the weighted fusion result obtained from step S4.

[0063] Then, the system compares the comprehensive score with the scores of each evaluation dimension to identify the dimensions with lower scores.

[0064] According to the identified dimensions with lower scores, the system will generate specific design optimization suggestions. The optimization suggestions will propose reasonable improvement measures according to the requirements of each evaluation dimension, such as using more energy-efficient building materials in the design solution, increasing the use of renewable energy, and improving the functionality of the space layout.

[0065] As an option, the report can also provide solutions related to building design optimization, specifically including targeted optimization strategies according to different evaluation dimensions. For example, in the economic dimension, the report may provide suggestions to reduce costs by changing certain building materials or construction methods. In terms of user experience, the report may propose measures to improve user comfort by improving the space layout and enhancing the permeability of the building.

[0066] In a possible implementation, the system can also generate visual charts based on the comprehensive evaluation results, intuitively showing the scores of each evaluation dimension and their contributions to the comprehensive evaluation results. Through this visual form, designers can more intuitively understand the performance of each dimension and quickly locate the parts that need to be optimized.

[0067] In addition, the system can also customize the format of the evaluation report according to different user needs. For designers, the system may provide detailed technical suggestions and optimization strategies; for managers or decision-makers, the system may focus on presenting the comprehensive evaluation results and key optimization suggestions for quick decision-making.

[0068] Please refer to the appendix Figure 2 , an intelligent evaluation system for building design projects based on multi-dimensional indicators, including: A data acquisition module, which is used to acquire building design data from multiple sources, including BIM data, environmental impact data, user behavior data, project cost data, and building performance data; A data preprocessing module, which is connected to the data acquisition module and is used to clean and standardize the acquired building design data; An evaluation system construction module, which is connected to the data preprocessing module and is used to construct a multi-dimensional evaluation system, set evaluation dimensions and quantification indicators, and deeply fuse the evaluation results of different data sources through deep learning algorithms to generate a comprehensive evaluation result; A reinforcement learning adjustment module, which is connected to the evaluation system construction module and is used to dynamically adjust the weights of the evaluation dimensions, calculate the current weights of each evaluation dimension through reinforcement learning algorithms, and automatically optimize the weight allocation according to real-time feedback; A real-time feedback mechanism module, which is connected to the data acquisition module and is used to collect the operation data of the building project in real time, including energy efficiency data and user behavior data, and feedback it to the reinforcement learning adjustment module through Internet of Things devices to optimize the weights of the evaluation dimensions; A comprehensive evaluation report generation module, which is connected to the evaluation system construction module and is used to generate a comprehensive evaluation report based on the evaluation results processed by deep learning algorithms and provide design optimization suggestions; A data fusion module, which is connected to the evaluation system construction module and is used to perform weighted fusion on the evaluation results from different data sources and generate a final comprehensive evaluation score.

[0069] This system module is written based on the content of the method technical solution, and its technical solution details are the same, so they will not be elaborated here.

[0070] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent evaluation method for architectural design projects based on multi-dimensional indicators, characterized in that, It includes the following steps: S1. Collect building design data from multiple sources, including BIM data, environmental impact data, user behavior data, project cost data, and building performance data; S2. Preprocess the collected data of various types, including data cleaning, standardization, and mapping, to integrate data from different sources into a unified format; S3. Construct a multi-dimensional evaluation system, set evaluation dimensions and quantitative indicators, and perform deep fusion on multi-dimensional data through deep learning algorithms to generate a comprehensive evaluation result; S4. Dynamically adjust the weights of each evaluation dimension through reinforcement learning algorithms, optimize the weight allocation based on historical design data and real-time feedback, and collect the operation data of building projects in real time to automatically adjust the evaluation weights; S5. Generate a comprehensive evaluation report and provide design optimization suggestions to assist design decision-making.

2. The intelligent evaluation method for building design projects based on multi-dimensional indicators according to claim 1, characterized in that In the step S1, the BIM data includes construction parameters such as the spatial layout, structural form, material selection, and construction technology of the building design; The environmental impact data includes building energy efficiency, carbon emissions, climate impact, and resource consumption data, and real-time data collection is carried out through Internet of Things devices and building management systems.

3. The intelligent evaluation method for architectural design projects based on multi-dimensional indicators according to claim 1, wherein, In the step S2, data cleaning includes removing outliers, filling in missing data, and removing redundant information; The standardization process ensures that data from different sources can be compared on the same scale by converting data from different sources into a unified dimension, and the Z-score standardization method is used to normalize the data.

4. The intelligent evaluation method for architectural design projects based on multi-dimensional indicators according to claim 1, wherein, In the step S3, the evaluation dimensions include design functionality, sustainability, economy, user experience, and environmental impact, and each evaluation dimension is scored by setting corresponding quantitative indicators; The deep learning algorithm uses convolutional neural networks and recurrent neural networks to extract features from different types of data, generate a comprehensive evaluation result, and perform weighted fusion on the scores of each evaluation dimension.

5. The intelligent evaluation method for architectural design projects based on multi-dimensional indicators according to claim 1, characterized in that The process of the weighted fusion can be expressed by the following formula: ； Among them, is the comprehensive evaluation result; is the weight of the th data source; is the feature extracted from the th data source.

6. The intelligent evaluation method for architectural design projects based on multi-dimensional indicators according to claim 1, wherein, In the step S4, the reinforcement learning algorithm adopts the Q-learning algorithm, and dynamically adjusts the weight allocation of each evaluation dimension by calculating the current weight and corresponding score of each evaluation dimension; The real-time feedback mechanism includes real-time collection of the energy efficiency data and user behavior data of building projects, connecting through Internet of Things devices and building management systems, and adjusting the weights in real time to optimize the evaluation results.

7. The intelligent evaluation method for architectural design projects based on multi-dimensional indicators according to claim 6, characterized in that, The specific formula of the Q-learning algorithm is as follows: ； Among them, is the current state and action of the value; is the learning rate, which controls the learning speed of the model; is the current state and weight corresponding reward value; is the discount factor, indicating the importance of future rewards; in, the maximum value of all possible actions.

8. The intelligent evaluation method for architectural design projects based on multi-dimensional indicators according to claim 1, characterized in that In the step S5, the deep learning algorithm includes: A convolutional neural network, which is used to process building spatial layout data; A recurrent neural network, which is used to process time series data; Perform weighted fusion on the features extracted by the two to generate a comprehensive evaluation result.

9. The intelligent evaluation method for architectural design projects based on multi-dimensional indicators according to claim 1, wherein The weighted fusion for generating the comprehensive evaluation result is expressed by the following formula: ； Among them, is the final comprehensive evaluation result; is the weight of the th data source; is the evaluation result of the th data source.

10. An intelligent evaluation system for architectural design projects based on multi-dimensional indicators, according to the intelligent evaluation method for architectural design projects based on multi-dimensional indicators described in any one of claims 1-9, characterized in that, It includes: A data collection module, which is used to collect building design data from multiple sources, including BIM data, environmental impact data, user behavior data, project cost data, and building performance data; A data preprocessing module, which is connected to the data collection module and is used to clean and standardize the collected building design data; The evaluation system construction module, which is connected to the data preprocessing module, is used to construct a multi-dimensional evaluation system, set evaluation dimensions and quantification indicators, and deeply fuse the evaluation results of different data sources through deep learning algorithms to generate a comprehensive evaluation result; The reinforcement learning adjustment module, which is connected to the evaluation system construction module, is used to dynamically adjust the weights of evaluation dimensions, calculate the current weights of each evaluation dimension through reinforcement learning algorithms, and automatically optimize the weight allocation according to real-time feedback; The real-time feedback mechanism module, which is connected to the data acquisition module, is used to collect the operation data of construction projects in real time, including energy efficiency data and user behavior data, and feedback them to the reinforcement learning adjustment module through Internet of Things devices to optimize the weights of evaluation dimensions; The comprehensive evaluation report generation module, which is connected to the evaluation system construction module, is used to generate a comprehensive evaluation report based on the evaluation results processed by deep learning algorithms and provide design optimization suggestions; The data fusion module, which is connected to the evaluation system construction module, is used to perform weighted fusion on the evaluation results from different data sources and generate the final comprehensive evaluation score.

Images