Compressive strength and anti-chloride ion penetration assessment of geopolymer mortar merging PVA fiber and nano-SiO₂ using RBF–BP composite neural network

Abstract In this study, we investigated the mechanical properties and chloride ion permeation resistance of geopolymer mortars based on fly ash modified with nano-SiO 2 (NS) polyvinyl alcohol (PVA) fiber metakaolin (MK) at dose levels 0–1.2% for PVA 0–2.5% NS. The Levenberg–Marquardt (L–M) back propagation (BP) neural network, as well radial-based function (RBF) was used to predict compressive strength mortar different admixtures nanoparticles fiber, wherein electric flux value index performance. RBF–BP composite network constructed study nanoparticle-doped ground mortars. According experimental results model, mean square error (MSE) observed be 0.00071943, root (RMSE) 0.026822, absolute (MAE) thereby showing higher prediction accuracy, faster convergence, better fitting effect compared single BP RBF models. combined artificial providing a new method future assessment penetration merging fibers NS in experiments engineering studies.