Estimating Compressive Strength of Concrete Using Neural Electromagnetic Field Optimization.

Area of Science:

• Civil Engineering
• Materials Science
• Computational Intelligence

Background:

• Concrete compressive strength (CCS) is a critical mechanical property for concrete structures.
• Accurate prediction of CCS is essential for quality control and structural integrity.
• Existing prediction methods may lack efficiency or accuracy.

Purpose of the Study:

• To develop a novel and efficient method for predicting concrete compressive strength (CCS).
• To optimize an artificial neural network (ANN) using electromagnetic field optimization (EFO) for CCS prediction.
• To compare the performance of EFO with other optimization algorithms (WCA, SCA, CFOA).

Main Methods:

• An artificial neural network (ANN) model was developed for CCS prediction.
• The ANN was optimized using the electromagnetic field optimization (EFO) algorithm.
• EFO's performance was benchmarked against the water cycle algorithm (WCA), sine cosine algorithm (SCA), and cuttlefish optimization algorithm (CFOA).
• Key concrete parameters (cement, slag, fly ash, water, superplasticizer, aggregates, age) were used as inputs.

Main Results:

• The ANN optimized with EFO (ANN-EFO) achieved the lowest mean absolute error (5.6236) in CCS prediction.
• ANN-EFO demonstrated higher prediction accuracy compared to ANN-WCA (5.8363), ANN-CFOA (7.6538), and ANN-SCA (7.8248).
• The EFO algorithm was significantly faster than the other optimization strategies evaluated.

Conclusions:

• The hybrid ANN-EFO model is a highly efficient and reliable approach for the early prediction of concrete compressive strength.
• ANN-EFO offers a user-friendly, explainable, and explicit predictive formula for convenient CCS estimation.
• The study recommends ANN-EFO for practical applications requiring accurate and rapid CCS assessment.