Size and configuration syntheses of rigid-link mechanisms with multiple rotary actuators using the constraint force design method

Abstract

This study presents a new synthetic approach for path-generating or function-generating rigid-body mechanisms in the framework of a new hybrid genetic algorithm. In spite of some advantages of synthesis methods in designing rigid-body mechanisms, some inherent issues remain in determining optimal sizes and configurations of rigid links with multiple rotary actuators. To alleviate these difficulties and limitations, we improve our previous contribution, called the constraint force design method, by parameterizing and optimizing the existence of links, the x-y locations of joints modeled by unit masses (particles), and the kind of selection between rigid and string links, using binary, integer, and binary design variables, respectively. This new genotype coding system for GA for the integer and binary phenotypes makes it possible to use a smaller number of unit masses to synthesize manifold configurations of rigid-body mechanisms. In addition, the locations and types of rotary actuators are parameterized with additional integer design variables for the synthesis of realistic rigid-body mechanisms. Furthermore, for efficient optimization of GA, a new h-GA integrated with the Sequential Quadratic Programming (SQP) optimization algorithm is also developed to optimize the locations of joints. To demonstrate the validity of the present constraint force design method, several mechanism synthesis problems are solved.