Comparative Analysis of Optimization Techniques for Mechanical Design Problems : Sequential Quadratic Programming, Pattern Search, and Genetic Algorithms

Abstract

Optimization plays a critical role in mechanical design, aiming to enhance performance and efficiency while adhering to design constraints. This paper presents a comprehensive study of three prominent optimization techniques—SQP, Pattern Search, and Genetic (GA)—and their application to mechanical design problems. Specifically, the research focuses on the optimization of tension/compression spring design, pressure vessel design, and three-bar truss design. The study evaluates these methods based on convergence speed, accuracy, and robustness. The SQP is found to be highly efficient for smooth problems, delivering rapid convergence and precise solutions. In contrast, Pattern Search and Genetic Algorithms demonstrate greater versatility and robustness when dealing with complex, non-smooth problem landscapes. Pattern Search is effective in navigating design spaces with discontinuities or noisy functions, while Genetic Algorithms offer a powerful global search capability, particularly useful in avoiding local optima. The comparative analysis provides valuable insights into the strengths and limitations of each optimization technique, guiding engineers and researchers in selecting the most suitable approach for various mechanical design challenges. These findings underscore the importance of choosing the right optimization strategy to address the specific characteristics of the problem at hand.