Skip to content

Latest commit

 

History

History
38 lines (31 loc) · 3.3 KB

README.md

File metadata and controls

38 lines (31 loc) · 3.3 KB
Raw

Weight Reduction of a Speed Reducer using Modified Particle Swarm Optimization and Shuffled Frog Leaping Algorithm

Objectives

  • Identify the minimal cost function evaluation for the given constraints
  • Identify the optimal combination of algorithm specific parameters (in this case - 𝜔, c1, c2) for the problem

Problem Diagram

Results

  • The cost function evaluation obtained using modified PSO and SFLA were 8.1790% lower than the result obtained using the Taguchi method by Ku et al. (1998), 1.1619% (Crude Monte-Carlo method) and 13.41% (Stray Process) lower than that obtained by Jan Golinski (1970), and around 11% lower than results obtained by other researchers.
  • Optimal parameter combinations:
    • PSO: 𝜔 = 0.5, c1 = 1.5 and c2 = 1.5, for a population size of 30
    • SFLA: 𝜔 = 1, c1 = 2 and c2 = 0.5, for a population size of 150 (15 memeplexes with 10 frogs each)
  • Modified SFLA generally takes lesser number of convergence iterations (CI) on average as compared to modified PSO.
  • The average run time (RT) for the same number of iterations (1000) is greater for modified SFLA.

Tools Used

  • MATLAB - A programming and numeric computing platform
    • To implement PSO and SFLA algorithms
  • Python - A high-level, general-purpose programming language
    • To statistically analyse project results
    • NumPy - A fundamental package for scientific computing in Python
    • Matplotlib - A comprehensive library for creating static, animated, and interactive visualizations in Python
    • Pandas - a fast, powerful, flexible and easy to use open source data analysis and manipulation tool
    • Seaborn - a Python data visualization library based on matplotlib

Other Resources Used

References