A novel intelligent approach mixing optimally diesel and animal fat recycled biodiesel for fuel creation

Abstract

This paper proposes a novel AI-based approach to finding efficient biodiesel mixtures, for fuel production from recycled animal fats. Animal fat wastes produce biodiesel with high-quality physicochemical properties and elevate the features of the new fuel that obtain enhanced oxidative stability and increased cetane number. An evolutionary computation approach based on genetic algorithms has been implemented to find feasible solutions corresponding to optimal mixtures between diesel and animal fat waste-based biodiesel. Evolutionary computation approaches are particularly capable of finding high-quality solutions to complex optimization problems. A special function for reckoning optimal fuel properties is defined in the paper, named the Total Evaluation Function Value (TMFV). This function blends the parameters of (a) fuel cost and (b) fuel density. The proposed approach gives to the researchers the ability to focus on the criterion that they believe is more important for fuel creation. Regarding the Genetic Algorithm, a new Operator Percentage Adjustment (OPA) mechanism for both Crossover and Mutation processes has been implemented for efficiency improvement in searching for optimal solutions. Especially in those cases where no improvement in the TMFV criterion value is observed, the mechanism is capable of amplifying the optimal solution detection process by increasing the number of solutions acquired through the mutation operation. Optimal fuel solutions have been produced experimentally at an Environmental Technology Laboratory, revealing that the new biodiesel fuel costs less than the common diesel (improvement ranges between 17.22% and 17.32% per experiment set), providing competitive fuel prices, while its consumption is expected to reduce pollutant emissions by having lower sulfur content. Note that this study succeeds in obtaining very small to zero deviations between computer-based experimental results and results acquired from the real fuel production process in the Laboratory (the latter results are available in the open information database in Mendeley.