首页    期刊浏览 2024年12月04日 星期三
登录注册

文章基本信息

  • 标题:Augmented Lagrangian Particle Swarm Optimization in Mechanism Design
  • 本地全文:下载
  • 作者:Kai SEDLACZEK ; Peter EBERHARD
  • 期刊名称:Journal of System Design and Dynamics
  • 电子版ISSN:1881-3046
  • 出版年度:2007
  • 卷号:1
  • 期号:3
  • 页码:410-421
  • DOI:10.1299/jsdd.1.410
  • 出版社:The Japan Society of Mechanical Engineers
  • 摘要:

    The problem of optimizing nonlinear multibody systems is in general nonlinear and nonconvex. This is especially true for the dimensional synthesis process of rigid body mechanisms, where often only local solutions might be found with gradient-based optimization methods. An attractive alternative for solving such multimodal optimization problems is the Particle Swarm Optimization (PSO) algorithm. This stochastic solution technique allows a derivative-free search for a global solution without the need for any initial design. In this work, we present an extension to the basic PSO algorithm in order to solve the problem of dimensional synthesis with nonlinear equality and inequality constraints. It utilizes the Augmented Lagrange Multiplier Method in combination with an advanced non-stationary penalty function approach that does not rely on excessively large penalty factors for sufficiently accurate solutions. Although the PSO method is even able to solve nonsmooth and discrete problems, this augmented algorithm can additionally calculate accurate Lagrange multiplier estimates for differentiable formulations, which are helpful in the analysis process of the optimization results. We demonstrate this method and show its very promising applicability to the constrained dimensional synthesis process of rigid body mechanisms.

  • 关键词:Particle Swarm Optimization; Nonlinear Constraints; Mechanism Synthesis; Multibody Systems
国家哲学社会科学文献中心版权所有