Modified Differential Evolution for Task Assignment Problem

A modified Differential Evolution algorithm was proposed to solve the NP-complete Task Assignment problem. In the modified Differential Evolution algorithm, a new mapping operation was constructed to generate the permutation from the floating-point coded gene of each individual in the current population. This slightly modification made the extension of the continuous field of Differential Evolution to the discrete domain is possible. Two numerical experiments were conducted and the compared results with other heuristic algorithm show its superiority in solving the Task Assignment problem.