{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,10,14]],"date-time":"2026-10-14T04:47:13Z","timestamp":1791953233338,"version":"4.3.4"},"reference-count":32,"publisher":"MDPI AG","issue":"19","license":[{"start":{"date-parts":[[2026,10,1]],"date-time":"2026-10-01T00:00:00Z","timestamp":1790812800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"The Natural Science Foundation of Xinjiang","award":["2025D01C89"],"award-info":[{"award-number":["2025D01C89"]}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["32560422"],"award-info":[{"award-number":["32560422"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"award":["32560422"],"award-info":[{"award-number":["32560422"]}],"id":[{"id":"https:\/\/ror.org\/01h0zpd94","id-type":"ROR","asserted-by":"publisher"}]},{"name":"The second batch of Tianshan Talent Cultivation Plan for Young Talent Support Project","award":["2023TSYCQNTJ0040"],"award-info":[{"award-number":["2023TSYCQNTJ0040"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Sensors"],"abstract":"<jats:p>Path planning for safflower filament harvesting is a critical component for ensuring the efficient operation of robotic harvesting systems. This paper proposes a path planning algorithm named SA-TD3-RRT, which integrates the Rapidly-exploring Random Tree (RRT), Twin Delayed Deep Deterministic Policy Gradient (TD3), and a spatial attention mechanism to achieve hierarchical planning of safflower filament harvesting paths. In the RRT component, techniques including adaptive step size, goal-biased sampling, and Douglas-Peucker path simplification are introduced. Specifically, the adaptive step size strategy converts the fixed step size into a dynamically adjustable extension step, effectively improving both the search real-time nature and path quality of the algorithm. The purely random sampling is replaced with goal-biased sampling, guiding the search tree growth toward the goal while suppressing redundant nodes, thereby reducing path planning time. Post-processing techniques such as Douglas-Peucker are employed to significantly reduce path complexity and improve path quality. In the reinforcement learning component, a spatial attention module is added, enabling dynamic perception of obstacle distribution. A spatial awareness term is introduced into the reward function to obtain more favorable spatial attention rewards. Additionally, the smoothed RRT path is injected into the experience replay buffer as expert demonstrations, realizing policy guidance based on imitation learning. Simulation results show that under favorable operating conditions, the proposed algorithm reduces the planned path length by 85.64%, 82.69%, and 92.55% compared to TD3, DDPG, and PPO algorithms, respectively, and shortens the path planning time by 17.68% relative to the original TD3 algorithm. In a three-dimensional environment, the planned path length is reduced by 8.29% compared to TD3, and the planning time is decreased by 81.24%, reaching 4.72 s. Although the current results are based on rigorous evaluation and demonstration in a high-fidelity simulation environment that abstracts from unstructured planting conditions, the proposed SA-TD3-RRT algorithm consistently achieved significant performance gains over the baseline. This simulation-based feasibility study provides a solid algorithmic foundation and key parameter benchmarks for future deployment on physical harvesting platforms, effectively bridging the gap between algorithmic innovation and practical agricultural robotics.<\/jats:p>","DOI":"10.3390\/s26196250","type":"journal-article","created":{"date-parts":[[2026,10,2]],"date-time":"2026-10-02T08:15:18Z","timestamp":1790928918000},"page":"6250","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["RRT-Guided and Spatial Attention-Based TD3 Deep Reinforcement Learning Method for Safflower Filament Harvesting Path Planning"],"prefix":"10.3390","volume":"26","author":[{"ORCID":"https:\/\/orcid.org\/0009-0009-2110-6393","authenticated-orcid":false,"given":"Bangbang","family":"Chen","sequence":"first","affiliation":[{"name":"School of Mechatronic Engineering, Xinjiang Institute of Technology, Aksu 843100, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xiangdong","family":"Liu","sequence":"additional","affiliation":[{"name":"School of Mechatronic Engineering, Xinjiang Institute of Technology, Aksu 843100, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Liqiang","family":"Wang","sequence":"additional","affiliation":[{"name":"School of Mechatronic Engineering, Xinjiang Institute of Technology, Aksu 843100, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0009-0003-3912-9311","authenticated-orcid":false,"given":"Baojian","family":"Ma","sequence":"additional","affiliation":[{"name":"School of Mechatronic Engineering, Xinjiang Institute of Technology, Aksu 843100, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kejia","family":"Zhuang","sequence":"additional","affiliation":[{"name":"School of Mechanical Electronic Engineering, Wuhan University of Technology, Wuhan 430070, China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1968","published-online":{"date-parts":[[2026,10,1]]},"reference":[{"key":"ref_1","unstructured":"National Pharmacopoeia Commission (2020). Pharmacopoeia of the People\u2019s Republic of China, China Medical Science and Technology Press."},{"key":"ref_2","first-page":"66","article-title":"Research progress on nutrients and efficacy of safflower seed oil","volume":"10","author":"Wang","year":"2024","journal-title":"China Oils Fats"},{"key":"ref_3","first-page":"110","article-title":"Yield grading and yield increasing measures of safflower in Xinjiang","volume":"63","author":"Meng","year":"2024","journal-title":"Hubei Agric. Sci."},{"key":"ref_4","first-page":"162","article-title":"Detecting safflower filaments using an improved YOLOv3 under complex environments","volume":"39","author":"Zhang","year":"2023","journal-title":"Trans. Chin. Soc. Agric. Eng."},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"100","DOI":"10.1109\/TSSC.1968.300136","article-title":"A formal basis for the heuristic determination of minimum cost paths","volume":"4","author":"Hart","year":"1968","journal-title":"IEEE Trans. Syst. Sci. Cybern."},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"18658","DOI":"10.1109\/ACCESS.2020.2968471","article-title":"Hybrid RRT: A semi-dual-tree RRT-based motion planner","volume":"8","author":"Mashayekhi","year":"2020","journal-title":"IEEE Access"},{"key":"ref_7","doi-asserted-by":"crossref","unstructured":"Kuffner, J.J., and LaValle, S.M. (2000, January 24\u201328). RRT-connect: An efficient approach to single-query path planning. Proceedings of the 2000 IEEE International Conference on Robotics and Automation (ICRA), San Francisco, CA, USA.","DOI":"10.1109\/ROBOT.2000.844730"},{"key":"ref_8","doi-asserted-by":"crossref","first-page":"28","DOI":"10.1109\/MCI.2006.329691","article-title":"Ant colony optimization","volume":"1","author":"Dorigo","year":"2007","journal-title":"IEEE Comput. Intell. Mag."},{"key":"ref_9","first-page":"18","article-title":"Genetic algorithm","volume":"53","author":"Mathew","year":"2012","journal-title":"Rep. Submitt. IIT Bombay"},{"key":"ref_10","unstructured":"Mnih, V., Kavukcuoglu, K., Silver, D., Graves, A., Antonoglou, I., Wierstra, D., and Riedmiller, M. (2013). Playing Atari with deep reinforcement learning. arXiv."},{"key":"ref_11","first-page":"2545","article-title":"A path planning algorithm based on improved experience replay strategy","volume":"40","author":"Li","year":"2025","journal-title":"Control Decis."},{"key":"ref_12","first-page":"250","article-title":"3D environment path planning based on improved deep reinforcement learning","volume":"38","author":"Feng","year":"2021","journal-title":"Comput. Appl. Softw."},{"key":"ref_13","first-page":"128","article-title":"Lotus Seed Picking Path Planning Based on the Improved Pointer Network","volume":"48","author":"Zheng","year":"2026","journal-title":"J. Agric. Mech. Res."},{"key":"ref_14","doi-asserted-by":"crossref","first-page":"100083","DOI":"10.1109\/ACCESS.2023.3311519","article-title":"Mobile service robot path planning using deep reinforcement learning","volume":"11","author":"Kumaar","year":"2023","journal-title":"IEEE Access"},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"3803","DOI":"10.1109\/TNNLS.2019.2899311","article-title":"Kinodynamic Motion Planning with Continuous-Time Q-Learning: An Online, Model-Free, and Safe Navigation Framework","volume":"30","author":"Kontoudis","year":"2019","journal-title":"IEEE Trans. Neural Netw. Learn. Syst."},{"key":"ref_16","doi-asserted-by":"crossref","unstructured":"Gao, J., Ye, W., Guo, J., and Li, Z. (2020). Deep reinforcement learning for indoor mobile robot path planning. Sensors, 20.","DOI":"10.3390\/s20195493"},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"5253","DOI":"10.1109\/TII.2021.3125447","article-title":"A general framework of motion planning for redundant robot manipulator based on deep reinforcement learning","volume":"18","author":"Li","year":"2021","journal-title":"IEEE Trans. Ind. Inform."},{"key":"ref_18","unstructured":"Florensa, C., Held, D., Wulfmeier, M., Zhang, M., and Abbeel, P. (2017). Reverse Curriculum Generation for Reinforcement Learning. arXiv."},{"key":"ref_19","doi-asserted-by":"crossref","first-page":"4298","DOI":"10.1109\/LRA.2019.2931199","article-title":"RL-RRT: Kinodynamic Motion Planning via Learning Reachability Estimators from RL Policies","volume":"4","author":"Chiang","year":"2019","journal-title":"IEEE Robot. Autom. Lett."},{"key":"ref_20","doi-asserted-by":"crossref","unstructured":"Cai, R., and Li, X. (2024). Path Planning Method for Manipulators Based on Improved Twin Delayed Deep Deterministic Policy Gradient and RRT*. Appl. Sci., 14.","DOI":"10.3390\/app14072765"},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"12269","DOI":"10.1038\/s41598-026-42615-8","article-title":"A fuzzy-TD3 hybrid reinforcement learning framework for robust trajectory tracking of the Mitsubishi RV-2AJ robotic arm","volume":"16","author":"Hazem","year":"2026","journal-title":"Sci. Rep."},{"key":"ref_22","doi-asserted-by":"crossref","first-page":"378","DOI":"10.1177\/02783640122067453","article-title":"Randomized kinodynamic planning","volume":"20","author":"LaValle","year":"2001","journal-title":"Int. J. Robot. Res."},{"key":"ref_23","first-page":"283","article-title":"Faster path planning based on improved RRT-Connect algorithm","volume":"65","author":"Wang","year":"2019","journal-title":"J. Wuhan Univ. (Nat. Sci. Ed.)"},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"112","DOI":"10.3138\/FM57-6770-U75U-7727","article-title":"Algorithms for the reduction of the number of points required to represent a digitized line or its caricature","volume":"10","author":"Douglas","year":"1973","journal-title":"Cartographica"},{"key":"ref_25","doi-asserted-by":"crossref","unstructured":"Geraerts, R., and Overmars, M.H. (2006, January 9\u201315). Creating high-quality roadmaps for motion planning in virtual environments. Proceedings of the 2006 IEEE\/RSJ International Conference on Intelligent Robots and Systems (IROS), Beijing, China.","DOI":"10.1109\/IROS.2006.282009"},{"key":"ref_26","unstructured":"Fujimoto, S., Hoof, H., and Meger, D. (2018, January 10\u201315). Addressing function approximation error in actor-critic methods. Proceedings of the 35th International Conference on Machine Learning (ICML), Stockholm, Sweden."},{"key":"ref_27","unstructured":"Lillicrap, T.P., Hunt, J.J., Pritzel, A., Heess, N., Erez, T., Tassa, Y., Silver, D., and Wierstra, D. (2019). Continuous control with deep reinforcement learning. arXiv."},{"key":"ref_28","first-page":"5998","article-title":"Attention is all you need","volume":"30","author":"Vaswani","year":"2017","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_29","first-page":"100092","article-title":"Gated attention for large language models: Non-linearity, sparsity, and attention-sink-free","volume":"38","author":"Qiu","year":"2026","journal-title":"Adv. Neural Inf. Process. Syst."},{"key":"ref_30","unstructured":"Schulman, J., Wolski, F., Dhariwal, P., Radford, A., and Klimov, O. (2017). Proximal policy optimization algorithms. arXiv."},{"key":"ref_31","unstructured":"Haarnoja, T., Zhou, A., Hartikainen, K., Tucker, G., Ha, S., Tan, J., Kumar, V., Zhu, H., Gupta, A., and Abbeel, P. (2018). Soft actor-critic algorithms and applications. arXiv."},{"key":"ref_32","doi-asserted-by":"crossref","first-page":"529","DOI":"10.1038\/nature14236","article-title":"Human-level control through deep reinforcement learning","volume":"518","author":"Mnih","year":"2015","journal-title":"Nature"}],"container-title":["Sensors"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/1424-8220\/26\/19\/6250\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,10,14]],"date-time":"2026-10-14T04:11:37Z","timestamp":1791951097000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/1424-8220\/26\/19\/6250"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,10,1]]},"references-count":32,"journal-issue":{"issue":"19","published-online":{"date-parts":[[2026,10]]}},"alternative-id":["s26196250"],"URL":"https:\/\/doi.org\/10.3390\/s26196250","relation":{},"ISSN":["1424-8220"],"issn-type":[{"value":"1424-8220","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,10,1]]}}}