Robot path planning method in rough terrain based on multi-objective crossover-mutation particle swarm optimization
摘要
Automatic path planning in complex three-dimensional environments is a key problem for robots when performing tasks effectively. However, existing path planning methods, especially the traditional particle swarm optimization (PSO), face key challenges when dealing with such problems, such as easily falling into local optimal solutions, lack of diversity, and insufficient precision. These problems limit the performance and reliability of path-planning algorithms in complex environments. To overcome these limitations, this paper proposes a new hybrid path planning algorithm by introducing a state factor-based inertia weight updating scheme and merging the crossover-mutation operator into the multi-objective particle swarm optimization (MOPSO), which is called MOSFCMPSO. Firstly, by setting adaptive tuning parameters in MOPSO, the probability of particles falling into the local optimal solution is reduced. Secondly, crossover and mutation operators are added to increase the diversity of the population to avoid the early precocious phenomenon. Finally, to evaluate the performance of the proposed MOSFCMPSO, comparisons are conducted with several other algorithms: standard MOPSO, linearly decreasing inertia weight MOPSO (LMOPSO), MOPSO based on state factor-based inertia weight updating scheme (MOSFPSO), multi-objective grey wolf optimizer (MOGWO), non-dominated sorting genetic algorithm II (NSGA-II), and K-means clustering based MOPSO (K-MOPSO) across three terrains of varying complexity. The simulation results demonstrate that MOSFCMPSO outperforms all other algorithms regarding search accuracy, population diversity, and optimization results. Specifically, when compared to the most competitive state-of-the-art algorithm, the proposed method achieves a reduction in path length of