Multi-strategy Enhanced Pollination Algorithm And Its Application

The Flower Pollination Algorithm(FPA)is a new heuristic optimization algorithm proposed to simulate the self-pollination and cross-pollination behavior of plant flowers.The algorithm's cross-pollination and self-pollination processes use global search behavior and local search behavior,and use random perturbation behavior to balance the two search strengths.The algorithm has a simple structure,few adjustable parameters,strong search ability,and is easy to understand and implement.Since its introduction,it has received extensive attention from scholars at home and abroad,and has succeeded in many complex practical engineering optimization problems.However,with the gradual progress of the research,it is found that the algorithm converges slowly in the later period and easily falls into a local optimum,which limits the application range of the algorithm.This thesis mainly analyzes and improves the shortcomings of the flower pollination algorithm,with the purpose of improving the overall performance of the flower pollination algorithm and expanding its practical application range.The research content mainly includes the following three aspects:(1)Combining the flower pollination algorithm with the wind-driven optimization algorithm,a wind-driven pollination algorithm(WDFPA)is proposed.The wind-driven strategy greatly improves the global search ability of the flower-pollination algorithm and the convergence speed of the algorithm.WDFPA was used for 29 standard test functions and 2 engineering examples,and compared with other improved versions of the pollination algorithm and 6classic intelligent optimization algorithms.The experimental results show the superior performance of WDFPA.(2)The butterfly foraging strategy was introduced,and a butterfly strategy flower pollination algorithm(BFPA)was proposed to increase the search space of the algorithm and increase the diversity of the population.The Beetle Antenna Search Strategy was introduced to enhance the algorithm's local search ability and improve the algorithm's performance.The precision of the optimization at the later stage of the execution makes the algorithm have better overall performance in the process.BFPA is applied to the problem of IIR filters with three different orders.The experimental results show the performance of BFPA in solving practical problems.(3)The flower pollination algorithm is applied to solve the color quantization problem of color images,and the color quantization of color images based on K-means clustering is used as the basis to achieve equivalent color replacement.The experimental results show that the flower pollination algorithm can improve the color quality of images and reduce the color the loss in the quantization process presents the best image quantization effect with the optimal number of colors.