An new evolutionary multi-objective optimization algorithm

We introduce a new, simple and efficient evolutionary algorithm to multiobjective optimization problem, which based on neighborhood and archived operation (NAGA). The innovations contain two main parts: neighborhood identify procedure to obtain Pareto optimal solutions from the population and neighborhood crowding procedure to maintain the diversity of Pareto optimal solutions previously found. The neighborhood identify procedure is composed of two steps, first to identify the locally nondominated solutions from the population and then to obtain the global nondominated solutions among the locally solutions. The neighborhood crowding is introduced to maintain a widely distributed set of Pareto solutions along the Pareto optimal front, which through implementing a comparison among the neighborhood bounds of new identified Pareto solutions and those of solutions in the archive. The winners, which are not in any ranges of the solutions in the archive, will be copied to the archive. A well-tuned fitness assignment method is structured to guide the population converging to the true Pareto optimal front. This method is pragmatic compromise between the computational simplicity and efficiency. Four nicely balanced test problems are provided to check the performance of the approach.