Answers are grounded in this paper and cite their sources.
The superior mechanical properties of UHPC make it a high-performance cementitious material, but its nonlinear composition-property relationship is too complex for conventional statistical modeling. This manuscript presents the hybrid approach for enhancing UHPC performance for strength prediction. The proposed hybrid strategy is the joint execution of both Lagrangian Neural Network (LNN) and Mexican Axolotl Optimisation (MAO), termed as LNN-MAO method. The main goal of proposed approach is to raise strength, improve wet packing density, reduce carbon emissions, and enhance performance of UHPC. The LNN approach predicts mechanical properties and particle packing density in damp conditions. The MAO approach is employed to optimise the LNN algorithm's parameters. The proposed strategy is evaluated using MATLAB and demonstrates superior results, achieving error of only 2%, flexural strength of 24MPa, compressive strength of 23MPa, and carbon emissions of just 688kg. Compared to existing models like Artificial Neural Networks (ANN), Recurrent Neural Networks (RNN), and Back-propagation Neural Networks (BPNN), the LNN-MAO approach achieves higher accuracy and lower environmental impact, establishing a new benchmark for UHPC strength prediction. These indicate that proposed strategy is not only more accurate but also significantly more eco-friendly, making it promising solution for sustainable high-performance concrete design.
No takes yet. Share an insight, caveat, or question.
Mohan et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: