Abstract
As a global-optimized and naturally inspired algorithm, particle swarm optimization (PSO) is characterized by its high quality and easy application in practical optimization problems. However, PSO has some obvious drawbacks, such as early convergence and slow convergence speed. Therefore, we introduced some appropriate improvements to PSO and proposed a novel chaotic PSO variant with arctangent acceleration coefficient (CPSO-AT). A total of 10 numerical optimization functions were employed to test the performance of the proposed CPSO-AT algorithm. Extensive contrast experiments were conducted to verify the effectiveness of the proposed methodology. The experimental results showed that the proposed CPSO-AT algorithm converges quickly and has better stability in numerical optimization problems compared with other PSO variants and other kinds of well-known optimal algorithms.
Highlights
With the development of artificial intelligence and the increasing computational demands of industrial design stimulated by continuously developing industry, various optimization algorithms have been increasingly studied and applied in academia and industry
Chaotic Particle Swarm Optimization (CPSO)-Acceleration Coefficients (AT) with other kinds of well-known optimization algorithms (MFO, krill herd (KH) and biogeography-based optimization (BBO)) using the same conditions as the first group, and the parameter settings and experimental results are presented in Tables 5–7, respectively
When k was 0, the performance of the CPSO-AT optimization was worse than that of the other algorithms; when k was 1, the CPSO-AT optimization was slightly better than the other algorithms; when k was 1, the CPSO-AT optimization was superior to the other algorithms and the effect was more obvious
Summary
With the development of artificial intelligence and the increasing computational demands of industrial design stimulated by continuously developing industry, various optimization algorithms have been increasingly studied and applied in academia and industry. In the past 20 years, scholars and industry specialists have presented many natural heuristic algorithms, like particle swarm optimization (PSO) [4,5], sine cosine algorithm (SCA) [6], ant colony optimization (ACO) [7], biogeography-based optimization (BBO) [8], grey wolf optimizer (GWO) [9], differential evolution (DE) [10], krill herd (KH) algorithm [11], moth flame optimization (MFO) [12], and the whale optimization algorithm (WOA) [13] The proposed operational zones and aggregated operational zones are powerful tools for reliable comparison of deterministic and stochastic global optimization algorithms This new testing methodology is a promising method that can be widely applied by researchers from optimization fields in the near future. For special circumstances, such as solving complex optimization problems with different test functions, the robustness, population diversity and the ability to balance local exploitation and global exploration is still insufficient
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
