Article Preview
Top1. Introduction
The swarm intelligence optimization algorithm has a long history and has been developed for decades. Common intelligent optimization algorithms include particle swarm algorithm (PSO) (Yada, 2018), artificial bee colony algorithm (ABC) (Xue, 2018), ant colony algorithm (ACO) (Engin, 2018), Fish Swarm Algorithm (FSA) (Neshat, 2014), Cuckoo Search Algorithm (CS) (Yang, 2014), Hunting Algorithm (HuS) (Oftadeh, 2010), Differential Evolution Algorithm (DE) (Li, 2019), etc. Evolutionary computing, inspired by biological evolution, aims to solve complex problems in less time (Nayyar, 2018). Metaheuristic algorithms have made great progress in recent decades, and the above algorithms are all proposed on this basis, which are used to optimize and find solutions to computer science problems (Nayyar, 2018). The current popular stochastic biological optimization techniques also include genetic algorithms, which perform well in solving multi-objective optimization problems (Nayyar, 2018). Ant colony algorithm is still an efficient method for solving discrete optimization problems (Nayyar, 2016).
The particle swarm optimization algorithm was first proposed by Dr. Eberhart and Dr. Kennedy in 1995. The algorithm has a simple structure, is easy to implement, and can also handle some highly complex problems. The algorithm simulates the bird's foraging behavior in order to find the best food source (global optimal solution). The entire flock of birds divides and cooperates, and while looking for things, the flocks transmit their position information to each other (Kennedy & Eberhart, 1995; Eberhart & Kennedy, 1995). At the end of the algorithm iteration, the whole flock of birds gathers near the food source. This phenomenon can be called that the flock of birds has found the optimal solution to the problem, and the problem converges at this position.
Once the particle swarm optimization algorithm was proposed, it has caused extensive influence in the academic circle. The reason is that the algorithm is simple in structure, easy to understand, easy to implement, and has few parameter settings, and it has been applied to many real-world problems. Such as function optimization problems (Chen, 2018), motion planning problems (Kim & Lee, 2015), resource allocation problems (Gong, 2012), image processing problems (Setayesh, 2013) and so on.