The study presented by the student aims to predict pile load tests from pile load test data and soil investigation tests, as they are considered expensive and time-consuming tests, as well as to simulate the (LOAD-SETTLEMENT) relationship for piles and compare the ANN model display with the experimental results of pile load tests. In addition to conducting an analysis Sensitivity to knowing which of the input variables has the greatest impact on the model’s output. In addition to using several statistical tests to determine the accuracy and reliability of the model. The discussion committee consisted of the following: -

-Prof. Dr. Adnan Jayed Zaidan, Chairman

- Prof. Dr. Muayad Abd Al-Jabbar Ahmed, member,

-Prof. Dr. Ahmed Saud Muhammad, member

- Prof. Dr. Khaled Rasim Mahmoud, member and supervisor.

The thesis included developing a predictive model consisting of the inputs and outputs of the model, determining the size of the data, determining the optimal network structure and improving it, and verifying the optimal performance of the model and conduct a sensitivity analysis.

The results of the study indicated that the model containing one hidden layer with four nodes is the most accurate model in prediction. In addition to the results of the statistical tests of the ANN model, we can conclude that the model is a reliable and accurate way to predict the average settlement of piles based on the input parameters. The model has a relatively low RMSE for both test and training data, which indicates that it can accurately predict average normalization values and explains a large portion of the variance in average normalization values. The correlation coefficient (R) and coefficient of determination (R2) are both high, indicating a strong positive relationship between the average predicted and actual settlement values. A comparison was also made between the curves for the “PREDECTET and MESURED” values and showed a good match. The results of the sensitivity analysis also indicate that the LOG (LAD) entry has the greatest influence on the model’s outputs.

 

 

The discussion was attended by the Dean of the College of Engineering, Prof. Dr. Ameer Abdel Rahman Hilal, the respected and Assistant Dean for Scientific Affairs and Higher Studies, Assist. Prof. Dr. Mohamed Abdel Ahmed, and Assistant Dean for Administrative Affairs. Head of the Civil Engineering Department, Assist. Prof Dr. The respected Ahmed Tariq Noman and a number of studies students in the department. Congratulations to the student for obtaining the certificate, and to the discussion committee, success and success. We wish him to complete the scientific consultatio