Dimensional Synthesis of a Four-Bar Mechanism by Hybridization of Two Meta-heuristic Algorithms DE-PSO

A new approach for the dimensional synthesis of a four-bar linkage is proposed. This approach is obtained by combining the differential evolution (DE) meta-heuristic algorithm with the particle swarm optimization (PSO) which are briefly explained. The hybridization of these 2 is done by High level Relay Hybridization where the final solutions of DE are used as initial variables of PSO. This approach can be used to resolve problems involving multiple objectives like minimizing error metrics in position. The objective function used for the approach takes in account the Grashof criterion constraint and is used to compare the mutants and particle positions (which are base vectors). The efficiency and accuracy is tested using the parameters and case proposed by Saeed Ebrahimi and Pedram Payvandy in "Efficient Constrained Synthesis of Path Generating Four-Bar Mechanism Based on the Heuristic Optimization Algorithms" and the synthesis errors are compared. The results show that our hybrid algorithm generates an optimization error of 0.010604 which is less than that observed in Saeed Ebrahimi and Pedram Payvandy’s work which are 0.2986 for PSO and 0.0151 for DE under the same conditions.