Some parameters directly affect the steam reforming and the determination of their optimal values will definitely produce an increment in the outcomes of the bio-hydrogen. Therefore, the main goal of the presented paper is to improve the bio-hydrogen production-based steam reforming using fuzzy modelling and modern optimization. Three input controlling parameters are varied to improve the output performance of the steam reforming process. These parameters are the reformer temperature (RT), steam-to-fuel ratio (SFR), and biodiesel-to-glycerol ratio (BGR). In the design of steam reforming of glycerol/biodiesel mixture, the experiments were carried out at predefined set of values for the three inputs. The RT (◦C), SFR and BGR (%) are assigned values of [500, 650, 800], and [0, 10, 20], respectively. The proposed strategy is composed of two sequential stages. First, by using the experimental data, a robust fuzzy model to simulate the output performance of steam reforming of glycerol and biodiesel fuel mixtures has been created. The results showed that the proposed model is superior to the RSM. Second, the optimal input controlling parameters are determined by marine predators algorithm (MPA). The results of the proposed MPA proved the robustness of the methodology in comparison to other recent optimizers. The results of the proposed strategy showed that the hydrogen production has increased by 5.74% and 4.8 % compared to the experimental and the RSM methodology, respectively. In conclusion, the findings of the proposed strategy illustrate the superiority in comparison to the experimental work and the RSM.