Genetic fuzzy rule extraction for optimised sizing and control of hybrid renewable energy hydrogen systems

A major challenge related to the design of a hybrid renewable energy hydrogen system is which energy sources to include and at what capacity, for regionally different potentials of renewable energy and hydrogen demand. In addition, once the plant is in operation, control variables need to be optimised. The problem resorts to an area of multiple criteria decision making referred to as multi-objective optimisation. The results obtained from these type of algorithms include not only one optimal solution, but a set of optimal solutions (Pareto front) thereby offering a system designer several options. This set of solutions can be hard to interpret and a method is needed to automatically extract useful design and control strategies from this information. A methodology that is quite successful in deriving human interpretable rules from this type of information is genetic fuzzy systems. In this work a k-means clustering algorithm is used to generate membership functions and a fuzzy rule-base is trained by means of a genetic algorithm. The genetic fuzzy system obtained is reduced by determining the minimum number of rules followed by a membership function reduction process. The reduced genetic fuzzy system is deemed more interpretable. Geographic weather data from three different sites are used to generate data to be used in the genetic fuzzy method. Results show that the technique provides valuable information that can be used for the design of such hybrid renewable energy hydrogen production systems.