Surendra, H.J.Deka, P.C.Rajakumara, H.N.2026-02-042022Soft Computing, 2022, 26, 21, pp. 11839-1184714327643https://doi.org/10.1007/s00500-022-06966-4https://idr.nitk.ac.in/handle/123456789/22354Artificial intelligence methods resemble human thinking structure that are used in hydrological modeling. In this work, water consumption estimation modeling is done using Mamdani fuzzy inference system. Different combinations of the models were developed by changing structures scenario such as: membership function, rules criteria, fuzzy set and defuzzification method. Mapping of input and output function are done using climatic variables and water consumption data. Rainfall, maximum temperature, minimum temperature and relative humidity were used as input factors and water consumption as output function. The reasoning mechanism of the fuzzy inference system calculates the recommended value of water consumption. Obtained value is compared with the actual recommended values to determine the usefulness of the system. The performances of the models were evaluated using performance indices such as correlation coefficient, mean square error and mean relative error. Results highlight that Mamdani fuzzy inference system is effective in actual application. © 2022, The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.Fuzzy inferenceFuzzy setsMean square errorWater supplyDefuzzificationsFuzzy inference systemsMamdani fuzzy inferencesMamdani modelMemberships functionModel-based OPCOutput functionsRecommended valuesTimes seriesWater consumptionMembership functions