Accuracy Optimizing Bipolar Fuzzy Fractional Programming Problems Using the Variable Neighborhood Search Algorithm

In this paper, we propose a new variable neighbourhood search algorithm for solving a bipolar fuzzy number fractional problem (FBVNS). The proposed algorithm utilises a modified version of Kerre's inequality, a tool employed to compare bipolar fuzzy numbers. The utilisation of the FBVNS algorithm enables the direct resolution of the bipolar fuzzy number fractional problem, obviating the necessity for conversion to a crisp number representation. The FBVNS algorithm is further enhanced by the introduction of a novel local search method, which is informed by ascending directions derived from the resolution of four distinct crisp mathematical programming problems. Finally, the efficacy of the proposed algorithm is demonstrated through its implementation on a range of problems.