Path Planning for Agricultural Robots Based on Improved Reinforcement Learning Algorithms
Year: 2026
Type: conference-paper
Abstract: When agricultural robots use reinforcement learning algorithms for path planning, they often face challenges such as difficulty in locating target points, slow algorithm convergence, and a tendency to get stuck in local optima. To address these issues, this paper proposes an improved reinforcement learning algorithm for agricultural robot path planning. First, the environment is modeled using a grid-based method. then, an adaptive function is introduced to balance the algorithm's global and local search capabilities, thereby improving the slow convergence rate observed in the early stages of the reinforcement learning algorithm and effectively preventing it from getting stuck in local optima. Finally, the algorithm was validated through simulations with varying obstacle densities and training iterations. The results demonstrate that the proposed algorithm converges significantly faster overall, and the probability of finding the optimal path increases with more training iterations, thereby verifying the validity and superiority of the proposed algorithm.
Authors: Wenna Ding, Siyang Liu, Qianhang Yu
Institution: Dalian Art College
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W7208760490