Record Details

Intermittent Reservoir Inflow Prediction with Stochastic and Genetic Programming Models

DSpace at IIT Bombay

View Archive Info
 
 
Field Value
 
Title Intermittent Reservoir Inflow Prediction with Stochastic and Genetic Programming Models
 
Creator Kote, A. S
Jothiprakash, V
 
Subject Auto regressive integrated moving average
genetic programming
seasonal
normal distribution
Pawana reservoir
Maharashtra
India.
 
Description Genetic programming (GP) technique has been evaluated by applying it to the observed time series of Pawana reservoir in Upper Bhima River Basin, Maharashtra, India. The resulting GP models are compared with conventional univariate auto regressive integrated moving average (ARIMA) models. GP-a data mining technique has been investigated for predicting seasonal (June to October) reservoir inflow with a monthly time step. The results of the study revealed that both GP and ARIMA models could not predict the future inflows in a better way because the observed series has not followed any particular distribution. Hence the observed time series was transformed into normal distribution using logarithmic transformation and found that both the modeling approaches predicted better. Goodness-of-fit measures, standard statistics, time series, and scatter plot were used to validate the model performance. Encouraging results indicated that the logarithmic transformed GP models resulted in better and reliable forecasts of high and low inflows (extreme) compared to stochastic models. It is also found that if the input data follows normal distribution then there is substantial improvement in the accuracy of the models.
 
Publisher Indian Water Resources Society
 
Date 2012-09-12T11:05:31Z
2012-09-12T11:05:31Z
2010
 
Type Article
 
Identifier Journal of Indian Water Resources Society, 30(4) 10-19
0970-6984
http://dspace.library.iitb.ac.in/jspui/handle/100/14408
 
Language English