Open AccessResearch of the particle swarm method in the problem of optimizing the mode of motion of the manipulator along one of the generalized coordinates
Author(s) -
Dmitry Mishchuk,
Yevhen Mishchuk,
Ievgenii Gorbatyuk
Publication year - 2021
Publication title -
gìrničì, budìvelʹnì, dorožnì ta melìorativnì mašini
Language(s) - English
Resource type - Journals
eISSN - 2709-6149
pISSN - 2312-6590
DOI - 10.32347/gbdmm2021.98.0201
Subject(s) - particle swarm optimization , control theory (sociology) , swarm behaviour , context (archaeology) , function (biology) , polynomial , mathematics , computer science , mathematical optimization , artificial intelligence , mathematical analysis , control (management) , paleontology , evolutionary biology , biology
The problems of optimizing the modes of movement of mechanical systems, in particular robots and manipulators, is relevant in the context of modern society. Algorithms for optimal movements of components of robots and manipulators allow to realize complex trajectories of movements of their working bodies with predicted energy consumption, positioning accuracy, speed. Finding optimal modes of motion is a complex task that requires accurate formulation of the optimization function, constraint equations and determination of optimal laws that would meet the criteria of the optimization problem. In this article, the classical particle swarm method for finding the optimal mode of motion of the manipulator boom at one of the generalized coordinates was analyzed. The target energy function is the "energy" of the accelerations of the mechanical system, and the search for the optimal law of displacement was carried out using a fourth-order polynomial.
The theoretical study showed that the method of particle swarm can be used to find the optimal laws of motion, but when working with this method it is necessary to modernize the algorithm for determining its components, including particle velocity and their correction factors.
In determining the optimal laws of motion of the manipulator by the swarm method, this study uses an approach where it is assumed that time is discrete, and the value of the objective function was determined only at the accepted sampling points of time.