Shahin Jalili

Cultural algorithms (CAs) are metaheuristics inspired by biocultural evolution theory and possess unique features compared with other evolutionary algorithms. In recent years, they have been employed effectively for engineering optimisation problems. However, the performance of CAs can be significantly affected by the influence functions under different types of constraints and objective functions. This study proposes a hyper‑cultural algorithm (HCA) that employs a high‑level heuristic (HLH), called the failure‑reward rate of the multi‑armed bandit (FRRMAB) algorithm, to determine the appropriate influence function for each individual in the population space during the optimisation process. To validate the efficiency of the HCA, the optimum design of a 968‑bar, double‑layer space structure is performed. The results of the comparison reveal HCA's capability to deliver higher‑quality designs for large‑scale structures. The observations in this study confirm that employing multiple strategies during the search process leads to better designs.

Summary: This paper presents optics inspired optimization (OIO) as a new optimization technique for optimum design of truss structures with continuous variables. In OIO, the physical laws that govern the image formation in spherical mirrors are artificially modeled to develop a new metaheuristic method. In OIO, the surface of the objective function to be minimized is considered as the reflecting wavy mirror consisting of peaks and valleys. In the reflecting surface, each peak is assumed to reflect as a convex mirror and each valley to reflect as a concave one. By assuming each solution candidate as an artificial object (or light point) that is artificially glittered ray is reflected back by the function surface, the artificial image (new solution) is formed based on the governing equations in physics of optics. Compared with the other existing optimization techniques, OIO seems to be a promising alternative approach for solving truss optimum design problems.