Abstract
The flower pollination algorithm is a new metaheuristic optimization technique that simulates the pollination behavior of flowers in nature. The global and local search processes of the algorithm are performed by simulating the self-pollination and cross-pollination of flowers. However, the conventional flower pollination algorithm has several limitations. To overcome the problem of slow convergence and prevent the algorithm from becoming stuck around local optimum, this paper describes an enhanced metaheuristic wind-driven flower pollination algorithm (WDFPA). Experiments are conducted using 29 benchmark test functions and two engineering design problems, and the proposed WDFPA is compared against other metaheuristic optimization algorithms and several classical optimization approaches. The results show that WDFPA achieves better performance than the conventional flower pollination algorithm, especially in high-dimensional optimization problems. The convergence speed and accuracy of WDFPA exhibit significant improvements over other metaheuristic algorithms in many of the test cases. Additionally, WDFPA produces optimal results for engineering design problems involving a welded beam and a spring structure.
Highlights
Traditional optimization algorithms are useful for solving simple continuous or linear problems, but are limited in terms of solving large-scale combinatorial optimization problems, there are often great limitations, such as low efficiency, high cost, and high energy consumption
1) TEST RESULTS USING HIGH-DIMENSIONAL UNIMODAL FUNCTIONS The experimental results in Table 4 indicate that wind-driven flower pollination algorithm (WDFPA) outperforms the other algorithms in terms of optimizing high-dimensional unimodal benchmark functions
With the exception of f8, the average value given by WDFPA is less than that of the other comparison algorithms, and its convergence accuracy is improved
Summary
Traditional optimization algorithms are useful for solving simple continuous or linear problems, but are limited in terms of solving large-scale combinatorial optimization problems, there are often great limitations, such as low efficiency, high cost, and high energy consumption. The accuracy of the solution often falls short of the requirements. For this reason, many scholars have begun to study other techniques, such as metaheuristic algorithms. Inspired by intelligent behavior and natural evolution, many intelligent optimization algorithms have been proposed for solving complex optimization problems [66]. Particle swarm optimization (PSO) [1] is based on the simulation of bird predation behavior in nature, genetic algorithms (GAs) [2]–[5] simulate the evolutionary process of inheritance, variation, and natural
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
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
