Scholarlink

This research introduces an advanced version of a genetic algorithm designed to tackle multi-objective optimization challenges. NSGA-II selects next-generation individuals using crowding distance and non-dominated sorting. This study makes several key contributions: firstly, it ensures the distribution of selected members on the PF by proposing a novel method to address the shortcomings of traditional crowding distance. Secondly, it introduces a hybrid mechanism combining chaotic maps and opposition-based learning (OBL) to diversify and generate the initial population. Thirdly, it presents a new approach to evaluating the cost function for selecting individuals for the next generation. Proposed algorithm benchmarked against MOPSO, MOICA, MOGWO, MOCSO, and NSGA-II. The outcomes illustrate that the suggested approach excels in obtaining a set of optimal solutions on the PF surface while maintaining appropriate dispersion to achieve the Pareto optimal front (POF) and solve engineering multi-objective problems.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A Modified NSGA-II Algorithm Using New Crowding Distance and Chaotic OBL Techniques Applied to Multi-Object Engineering Design Problems

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Nastaran Ahmadi Ramezanloo,
  • Yousef Sharafi,
  • Yashar Jebraeily

摘要

This research introduces an advanced version of a genetic algorithm designed to tackle multi-objective optimization challenges. NSGA-II selects next-generation individuals using crowding distance and non-dominated sorting. This study makes several key contributions: firstly, it ensures the distribution of selected members on the PF by proposing a novel method to address the shortcomings of traditional crowding distance. Secondly, it introduces a hybrid mechanism combining chaotic maps and opposition-based learning (OBL) to diversify and generate the initial population. Thirdly, it presents a new approach to evaluating the cost function for selecting individuals for the next generation. Proposed algorithm benchmarked against MOPSO, MOICA, MOGWO, MOCSO, and NSGA-II. The outcomes illustrate that the suggested approach excels in obtaining a set of optimal solutions on the PF surface while maintaining appropriate dispersion to achieve the Pareto optimal front (POF) and solve engineering multi-objective problems.