Using Differential Evolution for the Transportation Problem with Fuzzy Coefficients

This study introduces differential evolution to solve the transportation problem with fuzzy coefficients. Differential evolution has received increasing attention owing to its simplicity and effectiveness in solving numerical optimization problems. The fuzzy coefficients in the objective functions include fuzzy demands and fuzzy supplies, which are represented as fuzzy numbers. The signed-distance measurement is used for ranking fuzzy numbers in the evaluation and selection process of the differential evolution algorithm. The proposed approach is not only to simulate fuzzy numbers for representing fuzzy coefficients, but also to find the best solution to the problem. The numerical simulation results show that differential evolution is at most as efficient as genetic algorithms for solving the fuzzy transportation problem.