Abstract
This paper presents a method using a fuzzy neural system for finding the optimum release from reservoir so that the annual hydropower generation is maximized. Inflow uncertainties are taken into account implicitly. The training set is determined by solving the deterministic operation problem for a large number of flow sequences. Inflows are used as input features and releases as outputs in the fuzzy neural system. Fuzzy rules are created from the input-output data by recursively dividing the input space into a set of subspaces using fuzzy partitions. The resultant fuzzy system is used to determine the structure of the fuzzy neural network and to set its initial weights. The training of the neural network with back propagation is very fast.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
