Application of Mamdani Fuzzy Inference System for Runoff Prediction
Journal: International Journal of Science and Research (IJSR) (Vol.6, No. 5)Publication Date: 2017-05-05
Authors : Falguni Parekh;
Page : 1234-1238
Keywords : Runoff; SCS-CN; MFIS; Linguistic Variable;
Abstract
Watershed is the area covering all the land that contributes runoff water to a junction. There are various methods to find runoff, but runoff calculated by Soil Conservation Service Curve Number (SCS-CN) is used in the present study. After calculating runoff by SCS-CN method an effort is made to develop Runoff model by Mamdani Fuzzy Inference System (MFIS). In the present study daily rainfall data from 1992 to 2013 of monsoon season i. e. June to September, of four rain gauge station is used. The objectives of present study are to find runoff by using SCS-CN method and to develop MFIS to predict runoff. The output runoff by MFIS is compared with runoff by SCS-CN method. To develop model whole length of data is divided into two parts i. e. training and for validation. In 7030 data sets, 70 % data are used for training and 30 % data are used for validation and similarly for 6040 and 8020 data sets. Three models have been developed using above data sets i. e. model 1 which is having 9 Linguistic variables, Model 2 which is having 5 Linguistic variable and model 3 which is having 7 Linguistic variables. The best MFIS model is Model 1 based on performance indices. The RMSE of model 1 is 3.61mm and coefficient of determination (R2) is 0.9868 in training phase and 4.38mm and 0.9899 in validation phase respectively.
Other Latest Articles
- An Approach for Situation Aware IOT Service based on SOA
- Predication of Students Performance Using Data Mining
- Performance Analysis of AODV and AOMDV Routing Protocols for Reducing the Energy Consumption in MANET
- Some Properties of Mortar and Concrete Using Brick, Glass and Tile Waste as Partial Replacement of Cement
- Management of Hospital Information System through GIS: A Review
Last modified: 2021-06-30 18:55:25