Hybridization of Particle Swarm Optimization with adaptive genetic algorithm operators

Abstract

Particle Swarm Optimization (PSO) is a popular algorithm used extensively in continuous optimization. One of its well-known drawbacks is its propensity for premature convergence. Many techniques have been proposed for alleviating this problem. One of the alternative approaches is hybridization. Genetic Algorithms (GA) are one of the possible techniques used for hybridization. Most often, a mutation scheme is added to the PSO, but some applications of crossover have been added more recently. Some of these schemes use adaptive parameterization when applying the GA operators. In this work, adaptively parameterized mutation and crossover operators are combined with a PSO implementation individually and in combination to test the effectiveness of these additions. The results indicate that an adaptive approach with position factor is more effective for the proposed PSO hybrids. Compared to single PSO with adaptive inertia weight, all the PSO hybrids with adaptive probability have shown satisfactory performance in generating near-optimal solutions for all tested functions.

A new hybrid teaching–learning particle swarm optimization algorithm for synthesis of linkages to generate path

“…It can be applied for the form error evaluation, which contributes significantly to the production of mechanical components [29]. PSO can also be hybridized by combining with adaptive crossover and mutation rates [30] or by combining PSO with ant colony optimization (ACO) for convex and non-convex economic load dispatch (ELD) problem of a small scale thermal power system [31]. Also, a cooperative PSO (CPSO) is applied for solving function approximation and classification problems with improved accuracy [32].…”

The implementation frameworks of meta-heuristics hybridization with dynamic parameterization

“…The time-vary approach use iteration number as a main factor of formulation while adaptive approach relies more on the PSO performances indicators such as particle fitness, global fitness and particle position. Details about the time-vary and adaptive formulations used in this papers are described in [25][26] respectively.…”

“…The α is bounded within 0.1 times of the particle dimension. The algorithms have been previously tested on several benchmark functions [25][26] and the Vehicle Routing Problem with Time Windows (VRPTW) [28]. At this time, the interest has been coined to test the algorithms in Facility Layout Problem (FLP) [29].…”

Dynamic parameterization of the particle swarm optimization and genetic algorithm hybrids for vehicle routing problem with time window

“…Previous experiments [35] that tested the time-vary and adaptive approaches on a set of well-known benchmark functions have found that the effective approach for time- vary is linear decreasing [25,36] and for adaptive is ISA [20] as presented in the following Figs 3 and 4. The benchmark functions are Sphere denoted as f 1 , Rosenbrock as f 2 , Rastrigin as f 3 , Levy as f 4 , Griewank as f 5 and Ackley's functions as f 6 .…”