Article Preview
TopIntroduction
Intelligent logistics vehicles can deliver goods in the transportation system of the intelligent logistics ecology. In large quantities of goods transported, the implementation of land logistics vehicles' distribution efficacy is higher (Lo Storto & Evangelista, 2023). Therefore, the path-planning problem of intelligent logistics vehicles occupies a pivotal position in intelligent logistics ecological transportation. Using path-planning algorithms for start-to-end path optimization of vehicles has also become a hot topic in intelligent logistics. The Dijkstra algorithm (Sundarraj et al., 2023) and the A-Star algorithm (Zhang et al., 2023) are classical and feasible optimal path-planning algorithms. They can find the shortest distance between two points with no obstacles. However, they have partial time complexity and memory consumption. Later, by applying more efficient approximate algorithms, such as the ant colony optimization (ACO) algorithms (W. Wang et al., 2023), the genetic algorithm (GA) (Xu & Li, 2023), the annealing simulation algorithm (SA) (Venkateswaran et al., 2022), and so on, the NP (nondeterministic polynomial time) problems with complex paths are solved, which increases the algorithms' real-time resource utilization rates. The ant colony algorithm is a potent technique for conducting global searches and can converge to globally optimal solutions, even in highly complex search spaces. When dealing with a large-scale problem, it can effectively utilize the approximate exact solution approach to achieve an extremely close to optimal solution while staying within reasonable time and computing constraints. Since the ant colony algorithm is designed in a distributed and discrete form and is a simulation of the cooperative behavior of ants, the travel path of each ant in the team represents a potential solution (Yi et al., 2020). Moreover, each ant can influence the path choice of other ants by releasing pheromones. This distributed and discrete computing model is suitable for dealing with multiple traveling salesman problems (MTSP) to find optimal solutions (Changdar et al., 2023). It is precisely because the design of the ant colony algorithm is discrete that it is more suitable to combine the path environments of logistics vehicles with the grid method of spatial environment unitization. This method simulates the smallest moving unit of logistics vehicles in path planning in a discrete environment. The two-dimensional region created by this division is called a grid. The grid method is used as an auxiliary scheme to implement obstacle avoidance so that the research object can bypass obstacles and reach the target as soon as possible (Han, 2021; Mathiyalagan, 2010; Ajeil et al., 2020).
Sometimes, the ant colony algorithm's positive feedback mechanism has the potential to cause issues like slow convergence or falling into local optimization (Dorigo & Socha, 2018). The pheromone and heuristic rules of the ant colony algorithms are redefined in the improved ant colony algorithm proposed by Zhang and Zeng (as cited in Zhao et al., 2016; Zeng et al., 2016). However, the initial path planning will remain confused and fall into the local optimal solution. After that, although the EH-ACO algorithm of Gao et al. (2020) can reduce the probability of falling into local optimization and improve the efficiency of complex grid maps’ search, there will still be a distance deviation between the grid maps and the driving paths in the actual environments. The ant colony algorithm also has some limitations in avoiding dynamic and static obstacles in the driving environment with random interventions. It is hard to meet the actual environmental driving demands of logistics vehicles only by relying on the traditional ant colony algorithm combined with a grid map, which lacks natural environmental factors and has the shortcomings of avoiding obstacles.