Article Preview
Top1. Introduction
In nature, all kinds of organisms have their behavior strategie0d ways in the process of evolution. Inspired by these phenomena, people put forward many new methods and concepts to solve practical problems. The swarm intelligence optimization algorithm is an evolutionary algorithm of random search. The main idea is to simulate the foraging behavior of group creatures, such as fish schools, bird groups, and wolves. They will search for food in a cooperative way and constantly exchange food in the process. information to get more quality food as quickly as possible. Swarm intelligence has strong robustness, and the interacting individuals in the group are distributed, have no direct control center, and will not affect the solution of the problem due to the failure of a small number of individuals. The structure is simple and easy to implement, each individual can only perceive local information, and the rules that individuals follow are simple.Many classical algorithms such as particle swarm optimization (PSO) (Kennedy, James, and Russell C. Eberhart, 1997), Grey wolf optimization algorithm (GWO) (Mirjalili et al., 2014), ant colony algorithm (ACO)(Dorigo M et al., 2006), whale optimization algorithm (WOA) (Mirjalili S et al., 2016), and beetle antennae search algorithm(BAS)(Jiang X and Li S, 2017). They have been successfully applied in path planning (Wu Q et al., 2019), nonlinear control (Khan A H., 2019), image processing (Maitra M and Chatterjee A, 2008), and other fields.
The Sparrow Search Algorithm (SSA) is a new swarm intelligence optimization algorithm proposed in 2020 (Xue J and Shen B, 2020), Its principle is simple, the parameters are few, and the convergence speed is fast. It is more efficient than PSO, GWO, CO, and other algorithms in function optimization. Advantage. At present, SSA is also widely used in many practical engineering problems, such as vibration classification of rheostat transformers (Wu Y, 2021;Wang H and Xianyu J, 2021), flexible traction power supply systems (FTPSS) (Chen M et al., 2021), maximum power problems in the photovoltaic system(Zafar M H et al., 2021), the multi-objective problem of heater(Sukpancharoen S, 2021), prediction of water quality parameters in rivers (Song C et al., 2021), prediction of carbon price (Zhou J and Chen D, 2021;Zhou J and Wang S, 2021), Noise removal of measurement signals for concrete face rock fill dams (Xu L et al.2021), strength prediction of reinforced concrete(Li G et al.2021) bearing fault diagnosis (Xing Z et al., 2021), diabetes prediction (Wang Y and Tuo J, 2020).
However, it also has its shortcomings. For example, in the face of high-dimensional and complex problems, the optimization process always relies on a certain role, which reduces the learning ability of the algorithm and falls into a local optimum; on the other hand, there are more random parameters in the algorithm, resulting in the results being contingent.