Building Knowledge Graphs for Automated Renovation Scenarios
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Simulating single scenarios of green building renovations is manually intensive and dependent on manually specified characteristics, often leading to the neglect of critical building characteristics that could achieve impactful improvements.
Innovation Solution
A knowledge graph (KG) centric decision support system is employed to model building renovation scenarios, utilizing graph neural networks (GNNs) to recognize and recommend appropriate renovations based on building characteristics, and upload scenarios to a digital twin decision support system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual scenario simulation is used for green building renovations, then detailed analysis can be performed, but the process becomes manually intensive and time-consuming
Solution Approach 1:
The patent creates digital twins as virtual copies of physical buildings, allowing scenario simulations to be performed on the digital replica rather than requiring manual analysis of the actual building. This copying approach enables automated, rapid simulation while maintaining detailed analysis capabilities through the digital model.
Solution Approach 2:
The patent replaces manual mechanical analysis processes with automated computational systems. Machine learning models and algorithms automatically perform scenario simulations on digital twins, substituting human manual work with automated computational mechanics to reduce time while maintaining analysis depth.
2Ease of operation
If manual specification of building characteristics is used, then control over simulation parameters is maintained, but critical characteristics may be neglected
Solution Approach 1:
The patent implements feedback loops where the machine learning model analyzes building data, identifies critical characteristics that may have been overlooked, and automatically adjusts or supplements the simulation parameters. This feedback mechanism ensures comprehensive characteristic coverage while maintaining operational control through the digital twin framework.
Solution Approach 2:
The system performs self-service by automatically identifying and incorporating critical building characteristics through machine learning analysis. The digital twin autonomously captures relevant building attributes and simulation parameters without requiring complete manual specification, reducing the risk of neglected characteristics while maintaining ease of operation.
3Productivity
If comprehensive building characteristics are analyzed, then optimization potential is increased, but the complexity of scenario modeling increases
Solution Approach 1:
The patent segments the complex building renovation problem into manageable components by creating a digital twin that separates building characteristics, renovation scenarios, and performance metrics into distinct modular elements. This segmentation allows comprehensive analysis of multiple characteristics while reducing overall modeling complexity through structured decomposition.
Solution Approach 2:
The patent utilizes parameter changes in the digital twin model to represent different renovation scenarios and building characteristics. By systematically varying parameters within the digital twin framework, the system can analyze comprehensive optimization potentials while managing complexity through controlled parameter manipulation rather than complex structural modeling.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Knowledge graph (KG) centric decision support for GBN renovation scenario modeling includes populating a KG with different data pertaining to characteristics of different buildings in an GBN according to data type conforming to an ontology for the KG so that the KG associates different nodes corresponding to different ones of the data according linking relationships for the different nodes. KG enhancement queries are then executed against the KG to generate additional ones of the linking relationships. Yet further, different renovation scenario data structures are generated for one of the different buildings in the GBN by extracting from the KG different collections of interrelated characteristics for the one of the different buildings in the GBN and applying different modifications to one or more of the characteristics across different scenario data structures. Finally, one or more selected ones of the scenarios data structures uploads to a communicatively coupled digital twin decision support system.