Whale Optimization Algorithm Based on Dynamic Inertia Weight Factor

Aiming at the drawbacks of whale optimization algorithm (WOA) that the precision is poor, the convergence speed is slow, and it is easily trapped in local optimization when solving complex projects, this paper proposes an improved Whale Optimization Algorithm (IWOA).This strategy incorporates dynamic inertia weight factor into the updated expression of the traditional whale position to guarantee a counterbalance between the early and late algorithm’s ability to search for an optimal position, thus improving the search precision and rate of the WOA. In addition, for validating the effectiveness of our proposed algorithm in dealing with multidimensional and nonlinear optimization problems, six classical benchmark functions are selected as the test objective in this paper. The test functions contain unimodal and multimodal functions of fixed dimensions and are compared with genetic algorithm (GA), particle swarm algorithm (PSO) and standard whale optimization algorithm. The test results show that the improved algorithm converges faster and is more accurate than the original algorithm and several classical algorithms. And it has better results in the values of the two indicators, mean and standard deviation.

I. Introduction

It is characterized by the simplicity of the algorithms and ease of implementation of the meta-heuristic algorithms. They are efficient in avoiding falling into the local optimization trap. These algorithms have been evolved to rapidly address massive and complex projects to achieve a satisfying solution. They play a key role in a wide range of applications such as production scheduling [1], signal processing [2],image processing [3],and dynamic optimization [4]. The power of these algorithms lies in their ability to efficiently handle large and complex tasks, making them a powerful tool for solving the challenges of increasing complexity and data volume in today’s world. As a major concept of metaheuristic algorithms, group intelligence algorithms employ an iterative approach to continually update the population in search of an optimal solution. This approach highlights the synergistic role of the population in the algorithm, which enables the population intelligence algorithm to explore potential solutions in the problem space through the interaction and information exchange between individuals. Typical group intelligence algorithms involve genetic algorithm (GA) [5], which simulates Darwin’s theory of biological evolution and genetic mechanisms of biological evolution of nature evolution; and particle swarm optimization (PSO) [6], which emulates the bird population’s preying behavior.