A Cascaded Fuzzy Inference System for Indian river water quality prediction

Now-a-days, Fuzzy Inference System (FIS) is considered as an effective tool for solution of many complex engineering systems when ambiguity and uncertainly is associated with the systems. Mamdani and Takagi. Sugeno and Kang (TSK) models poses simplicity in modeling but their system performance prediction capability is severely affected as complexity of the problem increases. In a multi-input, multi-output situation where a system consists of many subsystems and different outputs are desired from each subsystem, an improved version of FIS must be adopted rather than developing FIS for each subsystem. When dealing with such a system, it is prudent to use cascading systems rather than developing models for individual systems. To this end, a new Cascaded Mamdani Fuzzy Inference System is proposed in this paper and its performance is evaluated with the help of prediction of Indian River water quality index (WQI). In general, WQI value is a dimensionless number ranging from 0 to 100 (best quality) and plays an important role in evaluating the water quality of rivers. The proposed model is designed to predict WQI for five rivers in India. The cascaded fuzzy system simplifies and speeds up the computation of WQI as compared to the currently existing standards. In this paper, the proposed model is compared with three International water quality criteria and it is found that the designed model results in accurate prediction. (C) 2011 Elsevier Ltd. All rights reserved.