Two-Level Quality Decision Support System for Building Structural Damage Prediction and Maintenance Solution Recommendation in the Operation and Maintenance Phase

Abstract

Traditional building quality management in the operation and maintenance (O&M) phase of the building life cycle has concentrated on inspection of quality defects by using new equipment or methods. However, there has been little research in terms of predicting a building’s structural quality problems and suggesting related feasible maintenance solutions. This study proposes a two-level building quality decision support system (DSS) using the hybrid feature selection and least-squares twin support vector machine algorithms to predict structural damage types and damage causes and suggest maintenance solutions for a building. The basic-level building quality DSS model, which can be regarded a predictive maintenance system, predicts the probable changes in structural damage types over time as a means of reminding building maintenance practitioners to pay close attention to the most probable damage requiring maintenance at different times. The inspection-level building quality DSS model, which can be treated as a proactive maintenance system, predicts the probabilities of structural damage causes and suggests related maintenance solutions. This research can further enhance the accuracy of structural damage cause analysis and establish valid maintenance solutions for inspectors. In addition, it can help to optimize the design and construction processes of buildings, which will further decrease structural quality defects in the O&M phase.