Parameterized Interior Floor Plan Design Based on Differentiable Renderer

He, Shuhan; Fu, Qiang; Li, Xueming

doi:10.1007/978-981-19-6226-4_14

Cited by 1 publication

(1 citation statement)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The optimization process was constrained by input boundaries, overlap, and weight constraints to ensure parameterization of the design and reasonable spatial partitioning. The effectiveness of this method in floor plan design was verified through ablation and control experiments [11].…”

Section: Introductionmentioning

confidence: 95%

Parametric Design of Office Furniture Partition Space Integrated with the Interactive Evolution Algorithm of FNT and Tree Structure

Chen,

Guan

2024

SCPE

View full text Add to dashboard Cite

Office furniture and its spatial layout design are playing an increasingly important role in improving work efficiency and employee comfort. However, the technology still faces some challenges. For instance, accurately simulating and evaluating the behavior and feelings of people in the office environment is difficult due to the high complexity of furniture spacing space design. It is important to address these issues. The study aims to explore the key technology and practical application of the parametric design of office furniture partition space based on the interactive evolution algorithm of tree structure. This paper proposes an improved version of the flexible neural tree model and corresponding algorithm. It also presents a design method based on the interactive differential evolution algorithm to optimize the automatic balance effect between global exploration and local development in the average shortening of the difference vector based on individual distribution. The results showed that all indexes were larger than or equal to other algorithms on 46 datasets. According to the Wilcoxon signed-rank test, the P-value was all less than 0.05, which is a significant advantage. Median, mean, and quartiles indicated that the overall performance of the algorithm was higher than the others. Furthermore, similarity evaluation-based Flexible Neural Tree algorithm had no outliers in the selected dataset, which also indicates the stability of the performance. The research results will support innovation and development in the field of office furniture design. This will promote intelligence, efficiency, and personalization in the design process, and meet the diverse needs of modern office environments.

show abstract

Section: Introductionmentioning

confidence: 95%