Use of neural network supervised learning to enhance the light environment adaptation ability and validity of Green BIM

ABSTRACT

This study proposes that a “predictive value” obtained through neural network learning be used instead of the “simulation value” in judging whether design goals have been met, and thereby enhance the optimization ability of Green BIM in the design decision-making process as a whole. There are inevitably discrepancies between Green BIM ‘s simulated performance data and the performance data obtained from the actual completed environment, neural network learning can be used in conjunction with training to obtain a predictive ability, and the resulting predictive values are more representative of actual performance than simulation values. In order to construct a simulated adaptive building façade based on light environment performance, this project plans to conduct the following six steps in a two-stage process:

Stage 1: data collection, learning algorithm, achieving predictive ability: (1) BIM modeling, (2) BPA performance simulation, (3) production of an actual structure and illuminance measurement, (4) and collection of sample data in order to perform training in supervised neural network learning.

Stage 2: After obtaining a predictive ability, finding an optimized proposal and implementing automated control: (5) Setting targets in order to find an optimized adaptation plan, and (6) implementation of script-oriented automatic control.

GRAPHICAL ABSTRACT

KEYWORDS:

• Green BIM
• neural network supervised learning
• CNS illuminance standards

Acknowledgement

The author wishes to acknowledge Architecture & Building Research Institute, Ministry of the Interior “Taiwan Green BIM Applications Framework Research” projects (case number: 10415B0007); Ministry of Science and Technology Green BIM-based Energy Conservation and Carbon Emissions Reduction Design Optimization Decision-making (MOST 104-2221-E-035-066)

ORCID

Shang-yuan Chen http://orcid.org/0000-0003-0249-5900