1. UKM Learning and Research Repository
2. PHD Thesis / Tesis PHD
3. Faculty of Engineering and Built Environment / Fakulti Kejuruteraan dan Alam Bina

Title:	Suspended sediment load prediction using multifunctional genetic algorithm-neural network
Authors:	Haitham Abdulmohsin Afan (P77768)
Supervisor:	Wan Hanna Melini Wan Mohtar, Dr.
Keywords:	Water resources engineering Hydraulic structures Universiti Kebangsaan Malaysia -- Dissertations
Issue Date:	26-Oct-2017
Description:	Suspended sediment transport prediction is a significant factor in water resources engineering as it affects directly the design of hydraulic structures and the management of water resources. Suspended sediment in water streams can act as a physical pollutant by increasing turbidity or as a chemical pollutant, being the primary carrier of adsorbed chemicals, particularly for finer sized particles. The estimation of erosion and transported sediment is accelerated due to mostly anthropogenic activities such as deforestation, poor agricultural practices, and massive development. The transportation of sediment to and through the river system consists of a number of complex phenomena including fluid-sediment interaction, and the characteristics of both flow and sediment should be taken into account. An accurate model for sediment prediction is a priority for all hydrological researchers. Many conventional methods have shown an inability to achieve an accurate prediction of suspended sediment. These methods are unable to understand the behavior of sediment transport in rivers due to the complexity, noise, non-stationarity, and dynamism of the sediment pattern. In the past two decades, Artificial Intelligence (AI) and computational approaches have become a remarkable tool for developing an accurate model. These approaches are considered a powerful tool for solving any non-linear model, as they can deal easily with a large number of data and sophisticated models. In this study, several AI models have been utilized to predict the suspended sediment load (SSL) from the antecedent values of SSL and water discharge in daily time scale. The multifunctional Genetic Algorithm-Neural Network (GA-NN) model is proposed and simulated using the suspended sediment load and flow discharge data at Johor River. The GA-NN model was compared with the traditional AI models (FFNN and RSM) and Global Harmony Search based Response Surface Model (GHS-RSM). The result shows that GA-NN has superior results (MAE=14.366), (RMSE= 24.560), (R²=0.911) than other models where a noticeable enhancement was observed on results by utilizing the GA in training ANN and input pattern selection. Therefore, the input selection task is an essential step in modeling for simplifying the mission for searching for the optimal solution in terms of achieving accurate prediction. In terms of examining the performance of GA in different climate zones and different river sizes, two AI (i.e. Hybrid and Parallel) models based on neural network trained by genetic algorithm was applied on the Mississippi and Kelantan Rivers. The proposed methods proved their ability to accomplish satisfactory results (R²=0.9108 for Kelantan) and (R²=0.963 for Mississippi) for the Parallel model. The comparison with previous work result shows the effectiveness of input selection on the model accuracy by enhancing the prediction more than 38% than the traditional model.,Ph.D.
Pages:	145
Publisher:	UKM, Bangi
Appears in Collections:	Faculty of Engineering and Built Environment / Fakulti Kejuruteraan dan Alam Bina

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.