Applying Differential Evolution to Aggregate Production Planning

Differential evolution (DE) has become a popular and promising technique in the recent decade. There are many successful applications implemented DE in scientific and engineering fields. However, we rarely found it applying to the field of highly constrained aggregate production planning (APP) problems. Besides, in the course of solving the APP problem, we discover that the most often used DE strategies have unsatisfactory performance to the problem. For alleviating the deficiency, in this paper, we propose an improved DE strategy and a constraint handling mechanism, called the winner-based constraint handling mechanism, for solving the highly constrained APP problem. In the computational study, some instances of the problem are experimented and analyzed to evaluate the performance of the approach with the most often used DE strategies. The experimental results show that the improved DE strategy with the winner-based constraint handling mechanism can provide particular performance for solving the APP problem than other DE strategies.