A Quick Scatter Search

The evolutionary approach called Scatter Search (SS) is successful in its theory and applications. However, SS has some shortcomings such as initial population scattered less evenly over the feasible solution space, the poor ability of finding global optima and the large number of objective function evaluations, so that we design an evolutionary algorithm called the Quick Scatter Search (QSS). QSS combines an orthogonal design method with SS to improve the shortcomings and satisfy the needs of practical projects. The experimental results prove that the proposed algorithm is superior to SS. Not only is QSS the number of its objective function evaluations less and the velocity of its convergence faster than SS, but also it is quicker to find global optimal solutions and even extremely close-to-optimal global solutions. Especially, the performance of QSS is better than that of SS in high dimension space.