Related Experiment Video
Updated: Jun 13, 2026

Image Recognition and Parameter Analysis of Concrete Vibration State Based on Support Vector Machine
Published on: January 5, 2024
Multi-objective optimization of low-noise pervious concrete using a stacking ensemble learning and NSGA-II approach
Fan Yu1,2, Haoyun Zhang1,2, SiYe Zhang3
1College of Civil Engineering and Architecture, China Three Gorges University, Yichang, 443002, China.
This study optimizes pervious concrete mix designs for low-noise pavements using Stacking ensemble learning and NSGA-II. The method balances acoustic, mechanical, and hydraulic properties, significantly improving sound absorption.
Area of Science:
- Materials Science
- Civil Engineering
- Acoustics
Background:
- Pervious concrete faces performance conflicts between acoustic, mechanical, and hydraulic properties, hindering its use in low-noise pavements.
- Optimizing mix proportions is crucial to balance these competing characteristics.
Purpose of the Study:
- To develop a multi-objective optimization method for pervious concrete mix proportions.
- To enhance sound absorption while maintaining permeability and compressive strength.
Main Methods:
- Established a comprehensive database for training a Stacking ensemble learning model for sound absorption prediction.
- Integrated sound absorption, compressive strength, and permeability models into the NSGA-II genetic algorithm for multi-objective optimization.
- Validated optimization results through experimental testing.
Main Results:
- Aggregate gradation significantly impacts sound absorption, with a 95.7% improvement in optimal vs. poorest gradations.
- The Stacking ensemble model achieved R²=0.97, surpassing individual models.
- The optimized mix proportion (O3) met permeability and strength standards while exceeding single-sized aggregate sound absorption.
Conclusions:
- The proposed optimization framework effectively resolves performance conflicts in pervious concrete.
- Optimized pervious concrete demonstrates superior sound absorption, permeability, and strength.
- The Stacking ensemble learning and NSGA-II approach provides a reliable method for material performance optimization.
Related Concept Videos
Design Example: Managing Concrete Workability
To address...
Effects of Air-entrainment in Concrete
Workability of Concrete
Concrete's workability is determined by its resistance to internal forces that arise when...
Segregation in Fresh Concrete
Reinforcements in Concrete
Preplaced Aggregate Concrete