Integrating rule-based policies into reinforcement learning promises to improve data efficiency and generalization in cooperative pursuit problems. However, most implementations do not properly distinguish the influence of neighboring robots in observation embedding or inter-robot interaction rules, leading to information loss and inefficient cooperation. This letter proposes a cooperative pursuit algorithm named Decentralized Adaptive COOperative Pursuit via Attention (DACOOP-A) by empowering reinforcement learning with artificial potential field and attention mechanisms. An attention-based framework is developed to emphasize important neighbors by concurrently integrating the learned attention scores into observation embedding and inter-robot interaction rules. A KL divergence regularization is introduced to alleviate the resultant learning stability issue. Improvements in data efficiency and generalization are demonstrated through numerical simulations. Extensive quantitative analyses are performed to illustrate the advantages of the proposed modules. Real-world experiments are performed to justify the feasibility of DACOOP-A in physical systems.

The authors regret that the original graphical abstract contained an error by accident. In particular, the values of the OH density for the calcined SBA-15 and solvent-&-Fenton-treated SBA-15 were swapped. These values contradict the main message of the study. The values contained in the paper are however correct. In summary, the paper's data are correct but the graphical abstract contained swapped data leading to a contradiction. Below, an amended graphical abstract can be found, as well as in the online journal website. The authors would like to apologise for any inconvenience caused. [Figure presented]