Scholarlink

<p>Modern truss constructions must be lightweight to avoid resonance Induced dynamic load failures. Classic gradient-based optimization techniques sometimes stagnate in local optima sets due to nonlinear, nonconvex frequency limitations. Metaheuristic methods are adaptable but lose diversity, producing early convergence and inefficient design space exploration. This study utilizes five analytically connected models (SS-PICE, CSFLA-SSM, CABC-KOR³, CICA ISR, and Auto-Optimization) to construct a unified chaotic optimization framework. SS-PICE forecasts modal frequencies with high accuracy using adaptive chaotic sampling, while CSFLA-SSM delivers frequency-safe feasible regions. CABC-KOR³ minimizes resonance amplification using Koopman-based spectrum dynamics, while CICA ISR offers robustness against material and loading uncertainty through interval-stochastic evaluations. Finally, ACE-BHT learns global convergence acceleration hyperparameters and chaos schedules. This sequential data-driven pipeline optimizes mass, spectrum safety margins, resonance amplification, and robustness. Benchmark truss system numerical simulations result in 11% mass reduction, 40–50% damping-response improvement, and over 70% computational cost reduction over chaotic and non-chaotic approaches. Paper presents a repeatable, physically interpretable, and globally convergent methodology for frequency-constrained truss optimizations.</p>

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

Chaotic frequency-constrained trusses optimization using chaotic metaheuristic algorithms

  • 期刊论文

  • 发表日期：

    2026年2月17日

查看全文
  • Sangita Meshram,
  • Gopal Malba Alapure,
  • Deepak G. B.,
  • Abhilash A. Sukhadeve,
  • Abhijeet Jaiswal,
  • Sagar Shelare

摘要

Modern truss constructions must be lightweight to avoid resonance Induced dynamic load failures. Classic gradient-based optimization techniques sometimes stagnate in local optima sets due to nonlinear, nonconvex frequency limitations. Metaheuristic methods are adaptable but lose diversity, producing early convergence and inefficient design space exploration. This study utilizes five analytically connected models (SS-PICE, CSFLA-SSM, CABC-KOR³, CICA ISR, and Auto-Optimization) to construct a unified chaotic optimization framework. SS-PICE forecasts modal frequencies with high accuracy using adaptive chaotic sampling, while CSFLA-SSM delivers frequency-safe feasible regions. CABC-KOR³ minimizes resonance amplification using Koopman-based spectrum dynamics, while CICA ISR offers robustness against material and loading uncertainty through interval-stochastic evaluations. Finally, ACE-BHT learns global convergence acceleration hyperparameters and chaos schedules. This sequential data-driven pipeline optimizes mass, spectrum safety margins, resonance amplification, and robustness. Benchmark truss system numerical simulations result in 11% mass reduction, 40–50% damping-response improvement, and over 70% computational cost reduction over chaotic and non-chaotic approaches. Paper presents a repeatable, physically interpretable, and globally convergent methodology for frequency-constrained truss optimizations.