To address the issue that traditional methods in path planning for artificial forest data collection robots struggle to balance optimal path length and computational efficiency, a multi-target path planning method based on crossover simulated annealing is proposed to enhance the intelligence level and operational efficiency of artificial forest data collection.

Firstly, the optimal path and distance between any two target radar nodes are calculated using the A* algorithm. Secondly, crossover operations from genetic algorithms are introduced to improve the way the traditional simulated annealing algorithm generates new solutions, allowing the algorithm to explore a larger solution space and find the optimal solution. Next, the two offspring solutions generated by the crossover operation are compared with the parent solutions, resulting in four main scenarios. Based on the solution quality and acceptance criteria, the acceptance standard for new solutions in the simulated annealing algorithm is further refined, accelerating the algorithm’s convergence. The improved simulated annealing algorithm is then used to generate the optimal visiting sequence of multi-target nodes. Finally, based on the optimal visiting sequence, the A* algorithm is used to connect the optimal paths to form the global closed-loop planned path.

Experiments were conducted using the TSPLIB dataset, and the results were compared with the simulated annealing algorithm. Experimental results show that, compared to the simulated annealing algorithm, the proposed method reduces the path length by 22.3% and shortens the runtime by 10.5%. Furthermore, the algorithm's performance was verified in an artificial forest data collection experimental scenario in the Olympic forest park north area in Haidian District, Beijing. The experiment shows that, compared to the traditional simulated annealing algorithm, the proposed improved algorithm further reduces the path length by 11.69% and shortens the time by 21.99%.

This study proposes a multi-target path planning method based on crossover simulated annealing, which improves the rationality, smoothness, and computational efficiency of path planning for artificial forest data collection robots. It provides technical support for precise monitoring, resource assessment, and intelligent management of artificial forests, offering valuable insights for the application of intelligent equipment in the field of forestry engineering.