In the field of hydrology, an
accurate and reliable flow forecasting is an important aspect for civil
engineers for a proper planning and management of a river basin which includes
flood control and warning, reservoir operation etc. The core state of the art
operational of a real-time flow forecasting system is a hydrological and
hydrodynamic simulation model which uses information of the current state of
river basin using precipitation, evapotranspiration, hydraulic model boundaries
to forecast dam water level, outflow from the dam. In case of reservoir
planning and management, various alternatives and parameters need to be
identified for achieving desired objectives. For a complex water resource
system including the multipurpose reservoir system, several conflicts may arise
in order to meet the demands. During monsoon period sometimes the water inflow
may exceed the reservoir storage capacity resulting in dam spillage and unprecedented
flood may occur in the downstream. On the contrary, during the lean period due
to low availability of water, the demand for downstream supply cannot be met which
might increase the low storage level of the reservoir. During the past decades,
a great number of algorithms and methods were developed which can be utilized
in the field of water resources system management. A number of models including
Artificial Neural Networks (ANN), Fuzzy Logic Networks, Evolutionary Computing
(EC), Genetic Algorithm (GA), Honey Bee Mating had a wide range of application
by the researchers in water resource planning and management. The selection of
the models might be perplexing by the decision makers. Thus, it is of utmost
importance to review on these computational intelligent models which are
associated with the reservoir operation and forecasting system. The aim of this
study to have a comparison among the different models and have a review along
with the difficulties which may arise during modeling techniques. Moreover, it
is important to know which model is more reliable where data availability is
scarce.