—This paper studies the accumulation of sediments in Bang Yai Canal, Phuket. The increase in sediment level decreases the water storage capacity of the canal which leads to shallow flooding problem. The study clustered the sediment into 3 levels using k-means algorithm, and then utilized an Artificial Neural Network to forecast the level of the sediment for the next three years using multilayer perceptron as a predictor. An efficient schedule for dredging the sediment in Bang Yai Canal can be constructed from the results of the study. Suspended sediments in tons per day from January 01, 2007 until December 31, 2011 were used as an input to our experiment. The result shows that the deposit of suspended sediments is peaked in October and is low in January of every year for the past five years. The level of sediments also increases every year for the next three years. We recommend that the municipality may alleviate the problem by dredging the sediments prior October, so that the canal drainage capacity is maintained.
—Data Mining, k-means algorithm, multilayer perceptron, sediment forecast.
The authors are with the Department of Computer Engineering, Faculty of Engineering, Prince of Songkla University, Kathu, Phuket, 83120, Thailand (e-mail: firstname.lastname@example.org, email@example.com).
Cite: Apichat Heednacram and Thammaratt Samitalampa, "Apichat Heednacram and Thammaratt Samitalampa," International Journal of Engineering and Technology vol. 6, no. 4, pp. 338-345, 2014.