A comparative study on Ant Colony Optimization algorithm approaches for solving multi-objective path planning problems in case of unmanned surface vehicles

The increased use of unmanned surface vehicles (USVs) to various applications in complex environments has led to the need of novel path planning approaches that can effectively address multi-objective path planning problems. Ant Colony Optimization (ACO) is a well-known and popular path search algorithm for graph-based problems that finds the shortest path. It has been widely used for solving path planning problems for robotic vehicles, and unmanned aircraft and surface vehicles, among others. To this end, this study focuses on extending the use of ACO to cope with multiple objectives by using fuzzy inference systems, such as Mamdani and Takagi-Sugeno-Kang (TSK), or the root mean square error (RMSE) criterion adopted to the state-of-the-art SIGPA algorithm. A comparative evaluation of these approaches is conducted in the context of solving different scenarios of multi-objective USV path planning problems. The results showed that ACO with Mamdani reached better performance in terms of solution quality compared to the other approaches under examination, while ACO with RMSE presented higher convergence speed and ACO with TSK balanced better among convergence speed and solution quality. Thus, each of the proposed approaches can be considered for multi-objective path planning of USVs depending the application needs.

Automated summary

This study aims to extend the application of Ant Colony Optimization (ACO) to multi-objective path planning problems by using fuzzy inference systems and the root mean square error criterion. Comparative evaluation of different approaches in solving multi-objective USV path planning problems showed that ACO with Mamdani achieved the best solution quality, ACO with RMSE had higher convergence speed, and ACO with TSK balanced better between convergence speed and solution quality.