Page 68 - tech fest 2025 ב״ש חוברת תקצירים

68
Geoengineering Characterization of Quarry Materials using Hyperspectral Imaging under Laboratory Conditions and Artificial Intelligence
EEE-A-13
Yona Daniel; yona195@gmail.com Adi Savadya; adi.savadya@gmail.com
Advisors: Dr. Isaac August1, Dr. Dagan Bakun Mazor1 1SCE - Shamoon College of Engineering, Be’er-Sheva
In this project, we explored the use of hyperspectral imaging in the near-infrared (NIR) range for rock characterization, focusing on evaluating the physical and mechanical properties of rocks. Traditional testing methods are often expensive, time-consuming, and unsuitable for field conditions. To address this challenge, we captured hyperspectral images of rock samples under controlled laboratory conditions, followed by radiometric calibration, preprocessing, and data filtering. After processing the data, we trained an artificial intelligence model to predict key properties such as density, water absorption, and compressive strength. This method enables rapid and nondestructive rock characterization without the need for physical sampling; it offers an efficient and cost-effective solution for engineering and geological applications.
Keywords: artificial intelligence, hyperspectral imaging, near-infrared (NIR), radiometric calibration, rock characterization
Design and Modeling of an Adaptive OFDM System for Optimal Power Allocation in a Rayleigh Fading Channel
EEE-B-01
Lior Ashurov; liorashurov1234@gmail.com Reut Mahadani; reut151120@gmail.com
Advisor: Dr. Moshe Zohar
SCE - Shamoon College of Engineering, Be’er-Sheva
Information transmission via wireless channels is challenging due to random signal attenuation, which impacts performance and increases the bit error rate (BER). Adaptive power loading (APL) can overcome this by dynamically distributing transmission power across subchannels, based on real-time channel conditions. This project designed and modeled an adaptive OFDM system using MATLAB, developing an algorithm for optimal power allocation in order to maintain constant error probability with minimal power consumption. We investigated system performance across various SNRs, error probabilities, and dictionary sizes, performing statistical analysis to characterize optimal usage and the trade-off between power, error, and dictionary size. Utilizing a Rayleigh fading channel, we maintained constant error probability while minimizing total power consumption, assuming conditions of ideal receiver feedback.
Keywords: adaptive OFDM, error probability, power allocation, Rayleigh fading, wireless communication