ENHANCING A NOVEL NEURAL NETWORK ALGORITHM FOR FORECASTING THE IDENTIFICATION OF SHAPES AND DEFECTS IN POLYMER CONCRETE PANELS

Abstract

The increased durability and performance features of polymer concrete panels have led to their widespread application in construction. The manufacturing of precise and effective techniques for identifying forms and flaws is vital to guarantee the high quality of these panels. To increase the accuracy of structure and defect-recognition in polymer concrete panels, this study presents a new Stochastic raven roosting optimization enhanced artificial neural network (SRRO-EANN) forecasting technique. The data sample used to assess the completed model fits the training dataset is referred to the test dataset. The Gaussian filter (GF) is a tool used in the pre-processing and Principal Component Analysis (PCA) feature extraction, leading to more effective utilization and understanding the defect capturing. The findings of the research indicate the effectiveness for the future development of forecasting technologies in the realm of quality control and building material inspection.

Keywords

• polymer concrete (pc)
• stochastic raven roosting optimization enhanced artificial neural network (srro-eann)
• concrete
• panel