Multiobjective Optimization Based on Coevolutionary Algorithm

With the intrinsic properties of multiobjective optimization problems in mind, multiobjective coevolutionary algorithm (MOCEA) is proposed. In MOCEA, a Pareto crossover operator, and 3 coevolutionary operators are designed for maintaining the population diversity and increasing the convergence rate. Moreover, a crowding distance is designed to reduce the size of the nondominated set. Experimental results demonstrate that MOCEA can find better solutions at a low computational cost. At the same time, the solutions found by MOCEA scatter uniformly over the entire Pareto front.