Estimating the Shear Strength of Binary Blended Concrete Incorporated with Hydrated Lime via Artificial Intelligence Technique | Chapter 6 | Engineering Research: Perspectives on Recent Advances Vol. 1

This analytical study examined the shear strength of blended Portland cement concrete with hydrated lime (HL) added as an additive. For a variety of mix ratios, 120 shear strength values were empirically acquired at 7, 14, 21, and 28 days. The ingredients of this concrete were water, portland cement (PC), HL, granite chips (GC), and river sand (RS). 96 of the findings were used to create a Levernberg-Marquardt backpropagation artificial neural network (ANN) for measuring the concrete's shear strength. The 24 outcomes that were not employed were utilized to test the forecast's efficacy of the ANN. The six input variables in the model were the proportions of water, curing age, PC, HL, RS, and GC. The measured value of the shear strength was the output variable. One hidden layer comprising 20 neurons was implemented. The highest 28-day shear strength value of 1.257 N/mm2 was recorded when 13.75% of PC was supplanted with HL for a water-to-cement ratio of 0.58. The ANN's performance demonstrated that the model was implemented well enough. The network forecast and experimental values yielded root mean square errors (RMSE) ranging from 0.0278 to 0.06536. These are nearly equal to zero. Furthermore, the calculated factor of agreement (IA) was found to be between 0.0475 and 0.1747. These are within the predetermined range of 0 to 1 for varied consistency. R-values for the training, validating, testing and for all data were 0.96978, 0.96303, 0.95739, and 0.96624 accordingly. These were all close to 1 meaning that the model fits very well with the data sets. The most significant average percentage error between the experimental outcomes and the forecasts made by the model was calculated to be 2.5066%. Finally, there is no longer a requirement for experimental laboratory study because the developed ANN can be utilized to forecast the shear strength of hydrated lime cement concrete with convincing accuracy thereby saving time and energy during the concrete mix design process.

 

Author (s) Details

 

C.T.G. Awodiji
Department of Civil and Environmental Engineering, University of Port-Harcourt, Nigeria.

 

D.O. Onwuka
Department of Civil Engineering, Federal University of Technology, Owerri, Nigeria.

S. Sule
Department of Civil and Environmental Engineering, University of Port-Harcourt, Nigeria.

 

Please see the book here:- https://doi.org/10.9734/bpi/erpra/v1/2978

Investigating the PCA Effect on the 3D Face Recognition System Speed | Chapter 7 | Technological Innovation in Engineering Research Vol. 5

In this study, we discuss leveraging 3-dimensional data to accelerate facial recognition. Compared to using 2-dimensional data, the usage of this data makes facial recognition safer since it is more challenging to replicate. Photos can be used to collect 2-dimensional data, but 3-dimensional data cannot be produced because it needs a third dimension—facial depth information—which can only be gathered from 3-dimensional objects. In contrast to most other research investigations, the three-dimensional (3D) face recognition procedure employed in this study utilised data immediately obtained from the Kinect Xbox camera system rather than going through the steps of the facial reconstruction process into 3D form. so that it can speed things up without lowering accuracy expectations. The backpropagation algorithm is directly fed data from the camera. This research just serves to demonstrate that the approach employed can speed up face recognition, hence this algorithm was chosen since it is easier to understand than CNN or other algorithms. The PCA approach is further employed in place of the reconstruction stage. With its ability to reduce the amount of computation required, PCA may speed up systems by simplifying the quantity of data. Two methods of testing are used. Backpropagation and PCA are both used in the first test, however Backpropagation is the sole method used in the second test. The method combining Backpropagation and PCA in conjunction increased speed by up to 34.2 times, but accuracy decreased by 8.5 percent, according to the results. In order to respond to the findings of this study, more research is required.

Author (s) Details:

Adhi Kusnadi,
Departemen of Informatic, Universitas Multimedia Nusantara, Tangerang, Indonesia.

. Wella,
Departemen of Informatic, Universitas Multimedia Nusantara, Tangerang, Indonesia.

Rangga Winantyo,
Departemen of Informatic, Universitas Multimedia Nusantara, Tangerang, Indonesia.

I. Z. Pane,
Departemen of Informatic, Universitas Multimedia Nusantara, Tangerang, Indonesia.

Please see the link here: https://stm.bookpi.org/TIER-V5/article/view/7466