Firefly Algorithm Based on Division of Labor for Solving Constrained Optimization Problems

The firefly algorithm is prone to premature convergence and poor diversity when dealing with constrained optimization problems, and to address this problem, this paper proposes a firefly algorithm based on division of labor for solving constrained optimization problems (FADL). The firefly population was first divided, and the individuals were sorted in order of fitness value from best to worst. The first half of the better firefly individuals is used for local exploitation, and the second half of the worse firefly individuals are used for global exploration, replacing the previous inferior solution with the better solution resulting from exploitation and exploration, thus effectively improving the diversity of the population; the addition of Lévy flights search mechanism on top of the division of labor effectively increases the population search range and improves the global traversal capability of the algorithm. In the experimental part, the algorithm is evaluated on 28 constrained optimization problems of CEC2017 benchmark, and the results show that it has higher convergence accuracy and faster convergence speed compared to the comparison algorithm FADL, which can effectively solve the constrained optimization problems.