Membership functions play a critical role in the performance of fuzzy logic controllers. However, the arbitrary nature of their definition is typically associated with inefficient results in modelling and control problems. This work describes a new approach to designing and optimizing FLCs, focused on the refinement of membership functions. The method consists of two interconnected stages. The first stage is aimed at simultaneous optimization of MFs and complex fuzzy control rules. At the core of the optimization process, there is the Walrus Tuna Whale Optimization Algorithm. This novel technique enables precise adjustment of MF parameters, including range and shape. In turn, the defined approach is characterized by flexible and highly efficient fine-tuning. In addition, the algorithm optimizes the choice of MF type in conjunction with fixed fuzzy control rules. In this way, a dual-optimization paradigm is established, and many combinations of MF types are investigated systematically. Furthermore, the performance of each such variant in the control system is evaluated meticulously. It continuously refines the FLC rules and MFs by applying the synergy of learning and optimization in both phases. The final results of experimentation provide sufficient evidence that the technique works accurately and helps reduce errors in complex control systems.