A Novel Fish Migration Optimization with QUasi-Affine TRansformation Evolutionary for Numerical Optimization and Application

Fish migration optimization (FMO) is an efficient swarm intelligence algorithm inspired by the migration process of fish. This paper proposes a new optimized structure-QUasi-Affine TRansformation Evolutionary for the Fish Migration Optimization (QTFMO). The new algorithm introduces the evolution matrix of the QUasi-Affine TRansformation Evolutionary (QUATRE) algorithm. Simultaneously, it also increases the disturbance of the individual optimal position to the search agent position in the fecundity stage. By improving the update of the strategy of individual positions, the particle movement process is more scientific and the search space is broader. It largely avoids premature convergence of FMO. We compared the new algorithm with five other intelligent algorithms through 23 different types of test functions and applied the new algorithm to the Capacitated Vehicle Routing Problem (CVRP). Experimental results reveal that the newly proposed QTFMO has powerful competitiveness and can achieve better path planning.