Abstract An earthen dam is a structure made of soil particles that are bonded together and compacted in layers using mechanical methods. depend on their weight to combat forces such as sliding & overturning. Seepage passage during earth dam is the principal cause of collapse owing to erosion, scouring, and piping. The passage of water during soil can result in the displacement of the particles. Ongoing motion induces erosion. This study depends on Artificial Neural Networks (ANN) for estimation seepage in Duhok dam, utilizing measured upstream water level and flow rate, as well as piezometric head measurements from four distinct parts of the dam structure. The findings indicated excellent model efficacy. Artificial Neural Network models necessitate less field data, rendering them advantageous for dam safety evaluations, providing insights into their relative efficacy in forecasting seepage in earthen dams under diverse scenarios. The results were derived from established statistical metrics R², MAE, MAPE and E-NASH. This research concluded that many ANNs has excellent forecasting capabilities.

Abstract Extreme-soft soil prevailed in the field, and preparing and testing such soil in the lab is not without physical challenges. Therefore, the present study aims to prepare extreme-soft soil using commercial bentonite and to treat such soil with different materials for evaluation of long-term performance. The extreme-soft soils, both untreated and treated, were experimentally evaluated by means of a modified vane undrained shear apparatus. Samples of extreme-soft soil are prepared from bentonite ranging from 2% to 10%, to reach this target. There is a separate treatment for the 10% prepared soft adding different lime percentages (up to 20%) and polymer (up to 5%). The soil samples from the treated samples were kept for curing period in open and closed-air conditions for up to 180 days. It has been observed in the study that untreated extreme-soft soil with untreated undrained shear strength between 0.001kPa to 0.0017 kPa correspondingly (bentonite 2% - 10%). Also, at 28 days curing time in a closed-air condition, the 20% lime-treated 10% bentonite extreme-soft soil had gained 1959% undrained shear strength. In addition, the extreme-soft soil treated with 10% bentonite and 5% polymer has 600% greater undrained shear strength than the open-air condition (28 days of curing). Also, at 28 days curing time of 180 days in a closed-air condition, the undrained shear strength of 10% lime-treated 10% bentonite extreme-soft soil has been increased by 1273%.

Abstract The recognition of emotion through Electroencephalogram data is essential for human computer interaction, mental health tracking, and emotion sensing computing. This paper integrates Recurrent Neural Networks (RNN) and a feature selection technique based on Particle Swarm Optimization (PSO) method within multiclass EEG classification paradigm. The method increases classification efficiency by feature selection from EEG signals which reduces the amount of computations needed in the first place. The span of emotion identification is aided by RNN model’s power in capturing the sequential dependencies of EEG signals. The experiments conducted demonstrated that the proposed solution performs better than traditional approaches in terms of accuracy and effectiveness. This opens avenues for the more precise and immediate applications of affective computing by providing a comprehensive solution for EEG emotion recognition.

Abstract A geological model was constructed using Petrel software for the Yamama Formation in the Siba oil field, southern Iraq. Information from four drilled wells in the area was used to define water saturation, porosity, and permeability. In addition, well-log data including sonic, density, neutron, gamma-ray, spontaneous potential (SP), and resistivity logs readings were used to determine the petrophysical properties of the formation. This work aims to study the occurrence of potential oil zones within the formation. Results indicate that zone 1 has low water saturation, high porosity, high permeability, and a high value of net to gross. For instance, the average porosity of the reservoir in Zone 1 is approximately 7.33%, compared to only 2% in Zone 2. Additionally, Zone 1 has a permeability value of 7 mD compared to only 0.2737 mD for Zone 2. Finally, Zone 1 has a water saturation value of 63%, while Zone 2 has a saturation of 89%. Based on the collective analysis of these factors, it was concluded that Zone 1 is more favorable for hydrocarbon development and production than Zone 2. Therefore, these findings can be valuable for optimizing well placement, production strategies, and overall reservoir development planning for oil resources.

Abstract Inadequate hole cleaning significantly impacts drilling operations. For example, poor wellbore cleaning can result in various drilling issues, including a reduced rate of penetration (ROP), early bit wear, and, in extreme situations, well loss due to stuck pipe. Numerous studies have been carried out to comprehend how to reduce transport efficiency and offer potential remedies for the issue. They frequently provide empirical correlations based on experimental data. Many engineering fields have recently adopted artificial intelligence and machine learning. Consequently, oil and gas companies have increasingly used artificial neural networks (ANN) to forecast a number of crucial metrics. The purpose of this study is to use artificial intelligence approaches to forecast hole-cleaning efficiency. Two layers, TANSIG and LOGSIG transfer functions, and several training functions were used to construct feed-forward backpropagation ANN models. A dataset of 1,620 experimental records served as the basis for the investigation. Cutting density and pressure losses are included in the input parameters. Additionally, the model input consisted of drilling characteristics such as the drill pipe rotating speed (RPM), flow rate (GPM), pipe, and hole inclination angle. The best-performing model was selected using a sensitivity analysis using 2, 4, 6, and 10 neurons for each transfer function (LOGSIG and TANSIG). With a correlation coefficient (R) greater than 0.9, the results showed that the constructed model accurately predicted the cutting transport efficiency (TE) in the wellbore. The findings demonstrated that as the number of neurons increases, the model's accuracy in terms of R for training and testing also increases. The expected and real TE utilizing the TANSIG transfer function, GDM versus GD learning function, and four different training functions indicate that using the GDM adaption learning function generally outperforms the GD function. For instance, the correlation coefficient (R) for 10 neurons using the GDM function was 97.45 compared to 97.08 for the GD function. Additionally, results indicate that the LOGSIG transfer function a bit overperforms the results estimated by the TANSIG function at two and four neurons. However, at higher numbers of neurons, the TANSIG function performs better. Therefore, it is recommended that using the TANSIG transfer function with the GDM be used for future predictions of transport efficiency (TE).