Improved shuffled frog leaping algorithm by using orthogonal experimental design

In Shuffled Frog Leaping Algorithm (SFLA), the worst frog' position is improved based on the experiences of the best local or global frog, in two steps separately. In this Algorithm, discovering more useful information of previous search experience through designing learning strategies is a challenging research. While each of these experiences may has better value on some dimensions, the other one may has better value due on some others. Hence, an orthogonal learning (OL) strategy as a learning strategy is proposed that combines good dimensions of them by orthogonal experimental design (OED). Combined dimensions' values form a more efficient guidance vector to guide the worst frogs leaping to global best area. This modified SFLA is introduced as Orthogonal Learning Shuffled Frog Leaping Algorithm (OLSFLA). The proposed strategy is evaluated on a set of 13 benchmark functions including unimodal and multimodal. Results confirm that this strategy in most of the time improves the performance of SFLA, offering faster global convergence, higher solution quality in comparison with some SFLA.