Resistance reduction method for building transmission and distribution systems based on an improved random forest model: A tee case study

Abstract

Building transmission and distribution systems account for more than 1/3 of building energy consumption. Local components are common in building transmission and distribution systems, and their resistance effect significantly increases energy consumption. The development of low-resistance local components is a beneficial way to solve this problem. Taking tees as an example, this paper proposes a resistance reduction method that uses an improved random forest model and provides dimensionless shape optimization parameters for tees with different area ratios. Through full-scale experiments, numerical simulations and energy dissipation analyses, the resistance reduction effects and mechanisms of the optimized tees are verified and analyzed. The results show that under different working conditions, optimized tees can simultaneously reduce the local resistance in the main line and branch line. The resistance reduction rate of the main line can reach 26 %–106 %, and the resistance reduction rate of the branch line can reach 9 %–145 %. Optimizing tees can significantly reduce energy dissipation in internal flow fields. Previous studies on resistance reduction have focused mainly on rectangular components that can be reduced to two dimensions and have adopted mostly a priori trial and error methods. This study proposes an a posteriori resistance reduction method for circular components, providing a reference for energy conservation and carbon reduction in buildings.