Article Preview
Top1. Introduction
Mobile robot path planning is one of the most basic and critical issues in the field of mobile robot research (Dewang et al., 2018; Mohanty & Parhi, 2013; Qu et al., 2008). Its purpose is to find a path with the shortest distance between the start point and the end point under the condition of known robot environment information and the path does not pass through any obstacles (Mo & Xu, 2015). The method of solving the path planning problem of mobile robots can be divided into two categories: traditional algorithms and intelligent algorithms. Traditional algorithms include artificial potential field method, fuzzy logic algorithm, viewable method, free space method, etc. (Mac Thi et al., 2016; Zhao et al., 2018). Since the path planning problem was proposed in the 1970s, these traditional algorithms have played an important role in the field of robot path planning and have achieved many research results. However, with the continuous expansion of mobile robot application fields, such as practical applications in marine science, industrial field and military operations, these traditional path planning optimization methods will have certain defects in dealing with these complex environments. For example, the artificial potential field method is easy to fall into a local minimum, and there is a problem of unreachable goals. The visualization method is very inefficient and cannot meet the real-time requirements of path planning. Fuzzy control algorithm is difficult to establish fuzzy rule base in complex and changeable environment and lacks intelligent obstacle avoidance strategy for dynamic obstacles. (Yu et al., 2019) In recent years, with the rise of artificial intelligence, more and more intelligent algorithms have been proposed and applied to the path planning optimization of mobile robots to overcome the limitations of traditional path planning algorithms. One of the important characteristics of these intelligent algorithms is that their operation mechanism is very similar to the biological group behavior or ecological mechanism in nature, and the efficiency of these intelligent algorithms is usually higher than that of traditional algorithms. The typical ones are genetic algorithm, ant colony algorithm, particle swarm optimization algorithm, artificial neural network algorithm, firefly algorithm, artificial bee colony algorithm, invasive weed algorithm and so on. Khaled Akka et al. (2018) proposed an improved ant colony algorithm to solve the robot path planning problem, using the stimulus probability to help ants select the next node, and using new pheromone update rules and dynamic adjustment of evaporation rate to accelerate the convergence speed and expand the search space. Long s et al. (2019) proposed an improved ant colony algorithm, which realized the efficient search ability of mobile robot in complex map path planning, and established the grid environment model. Faridi et al. (2018) proposed a multi-objective dynamic path planning method for multi robot based on improved artificial bee colony algorithm. This method improves the artificial bee colony algorithm and applies it to the neighborhood search path planner and the algorithm avoids falling into local optimum by adding appropriate parameters into the objective function. Kang Yuxiang et al. (2020) proposed an improved particle swarm optimization algorithm for robot path planning. According to the principle that variables in gradient descent method change along the negative gradient direction, an improved particle velocity update model is proposed. In order to improve the search efficiency and accuracy of particles, the adaptive particle position update coefficient is added. Zhang Yi et al. (2020) proposed an adaptive elite ant colony hybrid algorithm based on the lone wolf search mechanism. The visual area search mechanism of wolf is introduced in the elite ant colony algorithm and the adaptive enhancement function is designed to improve the ability of the ant colony to find the path in the elite ant colony algorithm search mechanism. Cao Xinliang et al. (2020) aimed at the defect of slow convergence speed in the process of robot path planning by ant colony algorithm, proposed to plan the robot global path planning based on the improved ant colony algorithm, delimited the optimal region in the grid map, and established a new initial pheromone concentration model to set the initial pheromone concentration of each point differently, so as to avoid the blindness of optimization and the convergence speed has been improved. Liu Yongjian et al. (2020) proposed an improved ant colony optimization algorithm. Through the establishment of grid map model, the state transition rule and roulette game method are used to select the next node, and the back off strategy is adopted for the ants trapped in deadlock to avoid falling into local optimum. At the same time, the algorithm improve the pheromone enhancement coefficient, improve the pheromone volatilization factor, establish the interlocking relationship between the pheromone factor and the required heuristic function factor, shorten the length of the shortest path, reduce the number of iterations and improve the convergence speed of the algorithm. Luo et al. (2019) used pseudo-random state transition rules to select paths, which improved the global search ability and convergence speed of the algorithm.