A data-driven method for prediction and design of biomass ash alkali-activated materials

This study proposes a data-driven framework based on XGBoost and the whale optimisation algorithm (WOA) for predicting and designing the biomass ash alkali-activated pure solid waste cementitious materials. Based on 112 sets of experimental data, an XGBoost model was developed to predict 3-day and 28-day compressive strength and flowability based on input parameters including microchemical composition, alkali activator concentration, and water-to-binder ratio (W/B). To optimise material performance, WOA was employed to design and refine the material mix ratios and alkali activator concentrations. The results indicated that the W/B is the most critical parameter influencing compressive strength and flowability. The model’s prediction errors remained within 5% for both the training and test sets, validating the accuracy and feasibility of the proposed method. The framework has broad applicability and can serve as a reference for the development of similar solid waste–based materials.