Industrial heritage landscape design optimization method based on big data analysis

Abstract

The invention relates to the technical field of industrial heritage protection and development, in particular to an industrial heritage landscape design optimization method based on big data analysis, which comprises the following steps: constructing a data acquisition framework, classifying and analyzing multi-dimensional data by adopting an improved weighted clustering algorithm, generating a preliminary design scheme, and dynamically optimizing by utilizing GIS (Geographic Information System) three-dimensional modeling and VR (Virtual Reality) technologies. And finally, comprehensively evaluating the scheme effect through a deep learning model and generating a sustainability score. According to the method, environmental characteristics, cultural inheritance and social demands can be comprehensively considered, the scientificity and accuracy of landscape design are improved, a design scheme with high sustainability and high public satisfaction is generated, and the method is suitable for landscape design optimization of various industrial heritage and has high practicability and popularization value.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of landscape design and big data analysis, and specifically relates to an industrial heritage landscape design optimization method based on big data analysis. BACKGROUND

[0002] With the acceleration of urbanization and the increasing awareness of industrial heritage protection, industrial heritage, as an important part of urban culture, its protection and reuse have become a research hotspot. In the process of protection and development of industrial heritage, not only the reinforcement of building structure and functional update need to be concerned, but also the scientificity and individuality of landscape design need to be considered comprehensively to meet the multiple needs of modern society for environment, culture and aesthetics. However, the existing industrial heritage transformation technology mainly focuses on building structure reinforcement, internal space function division and digital protection, and the application of big data analysis technology in landscape design optimization is relatively lacking, which leads to the lack of scientificity and accuracy of design decision, and it is difficult to fully meet the needs of modern industrial heritage protection and development.

[0003] Through retrieval, it is found that the patent with publication number CN106437190B proposes a reinforcement and transformation method for industrial building heritage, which realizes the reinforcement of building structure by retaining old outer walls and old concrete frame system, and completes the functional update and heritage protection by combining with space division and equipment design. However, this technical solution mainly focuses on the reinforcement of building structure and the functional division of internal space, and does not involve the optimization of landscape design, nor does it introduce big data analysis technology to support design decision. This makes it lack comprehensive consideration of environment, culture and social needs when dealing with complex industrial heritage landscape design, and it is difficult to realize the scientization and individualization of landscape design. In addition, the patent with publication number CN116805355B provides a multi-view stereo reconstruction method resistant to scene occlusion, which generates high-precision dense point cloud model by enhancing depth map with image contour information, providing technical support for digital protection of cultural heritage. However, this technical solution focuses on the application of digital reconstruction technology, mainly for generating three-dimensional model, and does not involve industrial heritage landscape design optimization based on big data analysis. Therefore, it has limitations in supporting the practical application of landscape design, and cannot effectively combine multi-dimensional information such as environmental data, social needs and cultural background, thereby limiting the innovation and practicality of design scheme.

[0004] The above problems show that the existing industrial heritage related technical solutions still have obvious deficiencies in landscape design optimization, especially in the application of big data analysis technology. The current technology is difficult to fully integrate multi-source data for scientific decision-making, leading to a lack of precision and innovation in landscape design in response to complex environments, cultural heritage and social needs. Therefore, an industrial heritage landscape design optimization method based on big data analysis is needed, which builds a data collection framework, deploys a sensor network to monitor environmental parameters (such as temperature and humidity, light) in real time, uses Internet crawlers to capture historical documents, social media comments and public feedback information, forms a multi-dimensional data set; use an improved weighted clustering algorithm to classify and analyze the data, mine the environmental characteristics of the industrial heritage area and the public's attention hotspots; based on the classification results, generate a preliminary landscape design scheme, and use geographic information systems (GIS) for three-dimensional modeling and effect simulation; introduce online questionnaire surveys and virtual reality (VR) technology to collect user feedback and dynamically optimize the design scheme; finally, use a deep learning model to evaluate the environmental adaptability, cultural heritage and public satisfaction of the scheme, and generate a sustainability score. This method can significantly improve the scientificity and precision of design decisions, meeting the comprehensive needs of industrial heritage protection and development in landscape design. SUMMARY

[0005] The present application belongs to the technical field of industrial heritage protection and development, and particularly relates to an industrial heritage landscape design optimization method based on big data analysis. In order to solve the problem of insufficient scientificity and precision of landscape design in the prior art, the present application provides a method for optimizing industrial heritage landscape design through multi-source data integration, dynamic feedback optimization and comprehensive evaluation. This method can fully consider environmental characteristics, cultural heritage and social needs, and generate a design scheme with strong sustainability and high public satisfaction. In the present application, an improved weighted clustering algorithm is used to classify and analyze multi-dimensional data, mine regional characteristics and public attention hotspots, combine geographic information systems (GIS) three-dimensional modeling and virtual reality (VR) technology to dynamically adjust the design scheme, and use a deep learning model to comprehensively evaluate the scheme effect, finally generating a sustainability score.

[0006] The industrial heritage landscape design optimization method based on big data analysis provided by the present application solves the above technical problems, and comprises the following steps: firstly, a data acquisition framework is constructed, a sensor network is deployed to monitor environmental parameters (such as temperature and humidity, light intensity) in real time, and historical literature, social media comments and public feedback information are crawled through an Internet crawler to form a multidimensional data set; secondly, an improved weighted clustering algorithm is used to classify and analyze the data, and the environmental characteristics and public attention hotspots of the industrial heritage area are mined; thirdly, a preliminary landscape design scheme is generated based on the classification results, and geographic information system (GIS) is used for three-dimensional modeling and effect simulation; fourthly, online questionnaire survey and virtual reality (VR) technology are introduced to collect user feedback, and the design scheme is dynamically optimized; and finally, a deep learning model is used to comprehensively evaluate the environmental adaptability, cultural heritage and public satisfaction of the scheme, and a sustainability score is generated.

[0007] In the present application, the improved weighted clustering algorithm improves the classification accuracy by introducing a weight coefficient adjustment rule, and the weight coefficient is dynamically adjusted according to the importance of the data type, and the formula is as follows:

[0008]

[0009] wherein (W i ) represents the weight coefficient of the (i)th data, (P i ) represents the environmental relevance of the data type, (Q i ) represents the social attention of the data type, and (α) and (β) are the adjustment parameters of the environmental relevance and the social attention, respectively, and the values are set to 0.6 and 0.4 by default through experimental verification.

[0010] Further, the present application uses a deep learning model to comprehensively evaluate the design scheme and generate a sustainability score, and the formula is as follows:

[0011] [S = λ1·E + λ2·C + λ3·U]

[0012] wherein (S) represents the sustainability score, (E) represents the environmental adaptability score, (C) represents the cultural heritage score, (U) represents the public satisfaction score, and (λ1), (λ2) and (λ3) are the weight coefficients of each score item, and the values are set to 0.4, 0.3 and 0.3 by default through actual case verification.

[0013] In the optimization scheme, the data acquisition framework includes a sensor network, an Internet crawler module, and a data cleaning module, ensuring diversified data sources and reliable quality. The sensor network is deployed at key nodes in the industrial heritage area, monitoring environmental parameters in real time and uploading them to a cloud database; the historical literature and social media data captured by the Internet crawler module are processed using natural language processing techniques to extract key information; and the data cleaning module eliminates redundant and noisy data to generate a standardized dataset.

[0014] In the further optimization scheme, the improved weighted clustering algorithm specifically implements the following steps: initializing the cluster center, calculating the distance of each data point to the cluster center and assigning a class label, updating the cluster center according to the weight coefficient adjustment rule, and repeating the iteration until convergence. Through the above steps, the algorithm can effectively mine the environmental characteristics and public attention hotspots of the industrial heritage area, providing a scientific basis for subsequent design.

[0015] In the GIS three-dimensional modeling and effect simulation process, layered modeling technology is used to construct terrain, building, and vegetation models of the industrial heritage area, and lighting rendering technology is used to simulate the landscape effect at different time periods. In the optimization scheme, the before-and-after comparison effect diagram is displayed in chart form, intuitively presenting the optimization effect of the design scheme. The online questionnaire survey and virtual reality (VR) technology can efficiently collect user feedback and dynamically adjust the design scheme. The online questionnaire survey covers three dimensions: environmental perception, cultural identity, and functional demand, while virtual reality (VR) technology allows users to intuitively experience the actual effect of the design scheme through immersive experience.

[0016] In the present application, through multi-source data integration, dynamic feedback optimization, and comprehensive evaluation, environmental characteristics, cultural heritage, and social needs can be considered comprehensively to generate a design scheme that is sustainable and highly satisfactory to the public. The present application is suitable for landscape design optimization of various industrial heritage and has strong practicality and promotional value. BRIEF DESCRIPTION OF DRAWINGS

[0017] Figure 1 : Data acquisition framework schematic diagram showing the composition and interrelationship of the sensor network, Internet crawler module, and data cleaning module.

[0018] Figure 2 : Improved weighted clustering algorithm flowchart detailing the specific steps of initializing the cluster center, assigning a class label, updating the cluster center, and iterative convergence.

[0019] Figure 3 : GIS three-dimensional modeling and effect simulation comparison chart presenting the terrain, building, and vegetation models and lighting rendering effects of the industrial heritage area before and after modeling.

[0020] Figure 4An online questionnaire and a virtual reality (VR) technology application diagram showing the actual effect of the user feedback collection process and the immersive experience design scheme. DETAILED DESCRIPTION

[0021] The present application provides an industrial heritage landscape design optimization method based on big data analysis, which realizes the improvement of the scientificity and accuracy of the design scheme through multi-source data integration, dynamic feedback optimization and comprehensive evaluation. The specific embodiments of the present application will be described in detail below with reference to the accompanying drawings.

[0022] First, constructing a data collection framework is one of the key steps in the implementation of the present application. As shown in Figure 1 , the data collection framework includes a sensor network, an Internet crawler module and a data cleaning module. The sensor network is deployed at key nodes in the industrial heritage area to monitor environmental parameters such as temperature and humidity, light intensity, etc. in real time. These sensors upload data to the cloud database through wireless communication technology to ensure real-time and reliable data. The Internet crawler module is responsible for scraping historical documents, social media comments and public feedback information. This module uses natural language processing techniques to extract key information such as cultural background descriptions, public attention hotspots and sentiment trends. The data cleaning module further eliminates redundant and noisy data to generate standardized datasets for subsequent analysis. For example, in a certain industrial site project, the sensor network covers the main building groups and green areas in the park, collects environmental data every hour, and forms a time series dataset through cloud storage. At the same time, the Internet crawler module scraped more than 100,000 comments from social media platforms, and after natural language processing, extracted high-frequency words and sentiment scores, providing an important basis for subsequent classification analysis.

[0023] Secondly, an improved weighted clustering algorithm is used to classify and analyze multi-dimensional data, and to mine the environmental characteristics and public attention hotspots of the industrial heritage area. As shown in Figure 2 , the specific implementation steps of the improved weighted clustering algorithm include initializing the cluster center, calculating the distance of each data point to the cluster center and assigning the class label, updating the cluster center according to the weight coefficient adjustment rule, and repeating the iteration until convergence. The weight coefficient adjustment rule is introduced in the present application to improve the classification accuracy, and its formula is as follows:

[0024]

[0025] where (W i ) represents the weight coefficient of the (i)th class of data, (P i ) represents the environmental relevance of the data type, and (Q i) represents the social attention of the data type, (α) and (β) are the adjustment parameters of environmental relevance and social attention, respectively. Their values ​​have been verified by experiments and the default values ​​are set to 0.6 and 0.4. In practical applications, it is assumed that an industrial heritage area contains multiple functional zones, such as an industrial heritage museum, a cultural and creative park, and an ecological park. By performing weighted cluster analysis on the environmental data and social attention of these zones, it is possible to find differences in the characteristics of different zones. For example, the social attention of the museum zone is higher, while the environmental relevance of the ecological park zone is more significant. This classification result provides a scientific basis for subsequent landscape design.

[0026] Then, a preliminary landscape design plan is generated based on the classification results, and a geographic information system (GIS) is used for three-dimensional modeling and effect simulation. Figure 3 As shown, the GIS 3D modeling process employs layered modeling techniques to construct models of the terrain, buildings, and vegetation within the industrial heritage area. Light rendering technology is then used to simulate landscape effects at different time periods. For example, in one industrial heritage project, designers first determined the thematic positioning of each functional zone based on the results of a classification analysis. For example, the museum area highlighted historical and cultural displays, while the ecological park area emphasized integration with the natural landscape. Next, a 3D model was constructed using GIS software. The terrain model was generated based on elevation data, the building model was completed by scanning existing structures and integrating them with design drawings, and the vegetation model was modeled based on plant species and distribution patterns. Light rendering technology was used to simulate landscape effects at different times of day, morning, noon, and evening, providing a visual representation of the design solution. Furthermore, before-and-after comparisons of the modeling results are presented in graphical form, clearly demonstrating the optimized design results.

[0027] Next, we introduced online questionnaires and virtual reality (VR) technology to collect user feedback and dynamically optimize the design plan. Figure 4 As shown in the figure, the online questionnaire survey covers three dimensions: environmental perception, cultural identity, and functional needs. The question design combines the results of classification analysis with the preliminary design plan to ensure that the survey content is targeted. For example, in a certain project, the questionnaire survey set questions related to the richness of the exhibition content in the museum area and the vegetation coverage rate in the ecological park area, and was distributed to the public through an online platform. At the same time, virtual reality (VR) technology allows users to intuitively feel the actual effects of the design plan through an immersive experience. For example, users can enter the virtual scene through VR equipment, observe details such as the internal layout of the museum and the design of the ecological park trails, and put forward improvement suggestions. The collected user feedback is organized into a data set as an important reference for dynamically optimizing the design plan.

[0028] Finally, a deep learning model is used to comprehensively evaluate the environmental adaptability, cultural heritage, and public satisfaction of the plan to generate a sustainability score. The comprehensive evaluation formula is as follows:

[0029] [S = λ1 E + λ2 C + λ3 U]

[0030] where (S) represents the sustainability score, (E) represents the environmental adaptability score, (C) represents the cultural heritage score, and (U) represents the public satisfaction score, (λ1), (λ2), and (λ3) are the weight coefficients of each score item, and their values are set to default values of 0.4, 0.3, and 0.3 through actual case verification. In practical applications, the deep learning model predicts the scores of the design scheme based on the training data set. For example, in a certain industrial site project, the model predicts the environmental adaptability score of the design scheme to be 85 points by analyzing environmental data, the cultural heritage score to be 90 points by analyzing user feedback, and the public satisfaction score to be 88 points. Finally, the sustainability score is calculated to be 87.5 points, indicating that the scheme has a high level of sustainability.

[0031] In summary, the present application improves the scientificity and accuracy of industrial heritage landscape design through data collection framework construction, improved weighted clustering algorithm classification analysis, GIS three-dimensional modeling and effect simulation, online questionnaire survey and virtual reality technology feedback collection, and deep learning model comprehensive evaluation. In practical applications, the present application is suitable for the optimization of landscape design of various industrial heritage, and has strong practicality and popularization value. For example, in a certain abandoned factory renovation project, the design scheme that takes into account environmental characteristics, cultural heritage and social needs is successfully generated by the above method, and has received high recognition and positive evaluation from the public.

Claims

1. A method for optimizing industrial heritage landscape design based on big data analysis, characterized by: The following steps are involved: A data collection framework was constructed, and a sensor network was deployed to monitor environmental parameters in real time. An internet crawler module was used to capture historical documents, social media comments, and public feedback to form a multidimensional data set. An improved weighted clustering algorithm was used to classify and analyze the multidimensional data, exploring the environmental characteristics of industrial heritage areas and public concerns. Based on the classification results, a preliminary landscape design plan was generated, and a geographic information system (GIS) was used for three-dimensional modeling and effect simulation. Online questionnaires and virtual reality (VR) technology were introduced to collect user feedback and dynamically optimize the design plan. A deep learning model was used to comprehensively evaluate the environmental adaptability, cultural heritage, and public satisfaction of the plan to generate a sustainability score.

2. The industrial heritage landscape design optimization method based on big data analysis according to claim 1 is characterized by: The data acquisition framework includes a sensor network (1), an Internet crawler module (2) and a data cleaning module (3). The sensor network (1) is deployed at key nodes in the industrial heritage area and is used to monitor environmental parameters in real time and upload data to a cloud database.

3. The industrial heritage landscape design optimization method based on big data analysis according to claim 1 or 2 is characterized by: The improved weighted clustering algorithm improves classification accuracy by adjusting the weight coefficient rule. The weight coefficient formula is: Among them, (W i ) represents the weight coefficient of the (i)th category data, (P i ) indicates the environmental dependency of the data type, (Q i ) represents the social attention of the data type, (α) and (β) are the adjustment parameters of environmental relevance and social attention, respectively.

4. The industrial heritage landscape design optimization method based on big data analysis according to claim 3 is characterized by: The specific implementation steps of the improved weighted clustering algorithm include initializing the cluster center, calculating the distance from each data point to the cluster center and assigning a category label, updating the cluster center according to the weight coefficient adjustment rule, and repeating the iteration until convergence.

5. The industrial heritage landscape design optimization method based on big data analysis according to claim 1 is characterized by: In the GIS 3D modeling process, layered modeling technology is used to construct the terrain, building and vegetation models of the industrial heritage area, and lighting rendering technology is used to simulate the landscape effects in different time periods.

6. The industrial heritage landscape design optimization method based on big data analysis according to claim 1 or 5 is characterized by: The online questionnaire survey covers three dimensions: environmental perception, cultural identity and functional requirements. Virtual reality (VR) technology allows users to intuitively feel the actual effects of the design plan through immersive experience.

7. The industrial heritage landscape design optimization method based on big data analysis according to claim 1 is characterized by: The formula for the comprehensive evaluation scheme of the deep learning model is: [S = λ1·E+λ2·C+λ3·U], where (S) represents the sustainability score, (E) represents the environmental adaptability score, (C) represents the cultural heritage score, (U) represents the public satisfaction score, and (λ1), (λ2) and (λ3) are the weight coefficients of each scoring item respectively.

8. The industrial heritage landscape design optimization method based on big data analysis according to claim 2 is characterized by: The historical documents and social media data captured by the Internet crawler module (2) are processed using natural language processing technology to extract key information, and the data cleaning module (3) removes redundant and noisy data to generate a standardized data set.

9. The industrial heritage landscape design optimization method based on big data analysis according to claim 1 is characterized by: The environmental parameters monitored by the sensor network (1) include temperature, humidity and light intensity, and the data are uploaded to a cloud database to form a time series data set.

10. The industrial heritage landscape design optimization method based on big data analysis according to claim 1 is characterized by: The default weight coefficients of the sustainability score are set to (λ1=0.4), (λ2=0.3), and (λ3=0.3).

Images