Parallel Whale Optimization Algorithm for Solving Constrained and Unconstrained Optimization Problems

TL;DR

This paper introduces a parallel version of the whale optimization algorithm (PWOA) that leverages multi-core processors to improve computational efficiency and search performance in solving both constrained and unconstrained optimization problems.

Contribution

The paper presents a novel parallel implementation of WOA that automatically utilizes available processors, enhancing speed and efficiency over the sequential version.

Findings

PWOA achieves better computational time than sequential WOA.

PWOA maintains the same solution quality as the sequential version.

Parallel metrics show significant speedup and efficiency improvements.

Abstract

Recently the engineering optimization problems require large computational demands and long solution time even on high multi-processors computational devices. In this paper, an OpenMP inspired parallel version of the whale optimization algorithm (PWOA) to obtain enhanced computational throughput and global search capability is presented. It automatically detects the number of available processors and divides the workload among them to accomplish the effective utilization of the available resources. PWOA is applied on twenty unconstrained optimization functions on multiple dimensions and five constrained optimization engineering functions. The proposed parallelism PWOA algorithms performance is evaluated using parallel metrics such as speedup, efficiency. The comparison illustrates that the proposed PWOA algorithm has obtained the same results while exceeding the sequential version in performance. Furthermore, PWOA algorithm in the term of computational time and speed of parallel metric was achieved better results over the sequential processing compared to the standard WOA.