Dimensional synthesis of a four-bar linkage mechanism via a PSO-based Cooperative Neural Network approach

Abstract

The dimensional synthesis problem is one of the challenging problems in robotics which has initiated several mathematical challenges. In this paper, a novel algorithm is proposed based on combination of Particle Swarm Optimization (PSO) and Cooperative Neural Network (CNN) for solving synthesis problem of a four-bar linkage which leads to an optimization problem. The cooperative network, so-called PS-CNN, consists of memory-retaining particles which collaborate together based on PSO algorithm in a cooperative interaction converge to the optimal dimensional synthesis solution. In the complete-connected network, each neuron provides a solution. Thereby, solutions are updated according to the neurons' memory, their interaction with other neurons and the global best solution of the neurons in order to provide a proper solution to the optimization problem. The objective of the optimization problem is to minimize the distance of the robot's end-effector from the 5 prescribed points by the user when traversing them. Simulation results reveal the desirable performance of the PS-CNN for robot synthesis with higher complexities. Furthermore, the proposed approach opens an avenue to extend it.