Prediction of Water Level Time Series Data for Dam at Selangor using Chaotic Approach and Local Linear Approximation Method

Abstract

The increase in demand of water resources is in line with the increasing number of residents especially in urban areas.This indicates that there is a need to predict water level so that an adequate water supply can be channelled to the main dam to meet the demands of the residents. Hence, this study is conducted in order to develop a prediction model using chaos approach for the time series data in Sungai Klang that is located in urban areas. Prediction based on chaos approach is divided into two phases which are; detection of chaos behaviour and prediction process. The detection of chaotic behaviour is performed by 0 – 1 test. Meanwhile, the prediction phase is conducted using the local linear approximation method (KPLS). The results show that the water level time series in Sungai Klang is chaotic using 0 – 1 test. Next, the prediction using KPLS shows that the prediction result is good since the predicted time series and the observed time series are almost the same and the value of correlation coefficient obtained is 0.8002. Therefore, the water level in Sungai Klang was successfully predicted using chaotic approach and it is expected to help those responsible authorities in optimizing water resources management in Selangor.