Optimal fuzzy-based power management for real time application in a hybrid generation system

This study presents a fuzzy-based optimal energy management scheme for a grid-tied hybrid generation system. The hybrid system under study includes a fuel cell, an electrolyser, and a hydrogen storage subsystem and is also capable of exchanging power with the local grid under hourly electricity pricing. The topic of energy management is presented in detail in the form of a non-linear constrained optimisation problem. A comprehensive mathematical formulation is applied to build an accurate model. Due to the complexity and large-scale nature of the problem, its algebraic model is given in general algebraic modelling system (GAMS). Having developed an off-line optimiser through interfacing GAMS and MATLAB, the optimal energy management problem is solved under different load profiles and the results are utilised to train a Sugeno-type fuzzy inference system that is responsible for real time energy management. The fine tuning of the fuzzy system parameters, mainly including the membership functions and the weighting coefficients, is made using subtractive data clustering. To verify the performance and validity of the proposed approach, the simulation results are presented and discussed in both off-line and on-line modes.