<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Publishing DTD v1.3 20210610//EN" "JATS-journalpublishing1-3.dtd">
<article article-type="research-article" dtd-version="1.3" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xml:lang="ru"><front><journal-meta><journal-id journal-id-type="publisher-id">novtexmech</journal-id><journal-title-group><journal-title xml:lang="ru">Мехатроника, автоматизация, управление</journal-title><trans-title-group xml:lang="en"><trans-title>Mekhatronika, Avtomatizatsiya, Upravlenie</trans-title></trans-title-group></journal-title-group><issn pub-type="ppub">1684-6427</issn><issn pub-type="epub">2619-1253</issn><publisher><publisher-name>Commercial Publisher «New Technologies»</publisher-name></publisher></journal-meta><article-meta><article-id pub-id-type="doi">10.17587/mau.25.520-529</article-id><article-id custom-type="elpub" pub-id-type="custom">novtexmech-1632</article-id><article-categories><subj-group subj-group-type="heading"><subject>Research Article</subject></subj-group><subj-group subj-group-type="section-heading" xml:lang="ru"><subject>РОБОТЫ, МЕХАТРОНИКА И РОБОТОТЕХНИЧЕСКИЕ СИСТЕМЫ</subject></subj-group><subj-group subj-group-type="section-heading" xml:lang="en"><subject>ROBOT, MECHATRONICS AND ROBOTIC SYSTEMS</subject></subj-group></article-categories><title-group><article-title>Планирование движения роботов в социальной среде через обучение с подкреплением</article-title><trans-title-group xml:lang="en"><trans-title>Planning the Movement of Robots in a Social Environment Via Reinforcement Learning</trans-title></trans-title-group></title-group><contrib-group><contrib contrib-type="author" corresp="yes"><name-alternatives><name name-style="eastern" xml:lang="ru"><surname>Станкевич</surname><given-names>Л. А.</given-names></name><name name-style="western" xml:lang="en"><surname>Stankevich</surname><given-names>L, A.</given-names></name></name-alternatives><bio xml:lang="ru"><p>Канд. тех. наук, доц., </p><p>Санкт-Петербург.</p></bio><bio xml:lang="en"><p>Stankevich L. A., Cand. of Tech. Sc., Associate Professor,</p><p>Saint-Petersburg.</p></bio><email xlink:type="simple">stankevich_lev@inbox.ru</email><xref ref-type="aff" rid="aff-1"/></contrib><contrib contrib-type="author" corresp="yes"><name-alternatives><name name-style="eastern" xml:lang="ru"><surname>Ларионов</surname><given-names>А. А.</given-names></name><name name-style="western" xml:lang="en"><surname>Larionov</surname><given-names>A. A.</given-names></name></name-alternatives><bio xml:lang="ru"><p>Техник-программист, </p><p>Санкт-Петербург.</p></bio><bio xml:lang="en"><p>Saint-Petersburg.</p></bio><email xlink:type="simple">larionov.aa-mail@yandex.ru</email><xref ref-type="aff" rid="aff-2"/></contrib></contrib-group><aff-alternatives id="aff-1"><aff xml:lang="ru"><institution>Санкт-Петербургский политехнический университет Петра Великого</institution><country>Россия</country></aff><aff xml:lang="en"><institution>The Great Peter Saint-Petersburg Polytechnic University</institution><country>Russian Federation</country></aff></aff-alternatives><aff-alternatives id="aff-2"><aff xml:lang="ru"><institution>ООО "Специальный технологический центр»</institution><country>Россия</country></aff><aff xml:lang="en"><institution>LTD "Special Technologic Center"</institution><country>Russian Federation</country></aff></aff-alternatives><pub-date pub-type="collection"><year>2024</year></pub-date><pub-date pub-type="epub"><day>11</day><month>10</month><year>2024</year></pub-date><volume>25</volume><issue>10</issue><fpage>520</fpage><lpage>529</lpage><permissions><copyright-statement>Copyright &amp;#x00A9; Commercial Publisher «New Technologies», 2024</copyright-statement><copyright-year>2024</copyright-year><copyright-holder xml:lang="ru">Commercial Publisher «New Technologies»</copyright-holder><copyright-holder xml:lang="en">Commercial Publisher «New Technologies»</copyright-holder><license xlink:href="https://mech.novtex.ru/jour/about/submissions#copyrightNotice" xlink:type="simple"><license-p>https://mech.novtex.ru/jour/about/submissions#copyrightNotice</license-p></license></permissions><self-uri xlink:href="https://mech.novtex.ru/jour/article/view/1632">https://mech.novtex.ru/jour/article/view/1632</self-uri><abstract><p>Обсуждается проблема управления движением роботов в социальной среде в местах скопления людей. Разработан и исследован алгоритм планирования движения мобильных роботов среди неподвижных и движущихся препятствий на основе обучения с подкреплением. В качестве прототипа выбран алгоритм GA3C-CADRL, в котором робот и препятствия рассматриваются как взаимодействующие агенты. Алгоритм был модифицирован и реализован с использованием рекуррентной нейронной сети LSTM для аппроксимации одновременно функции ценности и политики. Нейронная сеть обучалась на общем наборе данных, полученном путем обучения с подкреплением типа "актер—критик". Дополнительно разработаны компоненты rl_ planner и social_msgs для интегрирования предварительно обученного алгоритма планирования в систему управления роботом на программной платформе Robot Operating System 2. Первый компонент реализует обработку входных данных, вычисление действия робота и формирование требуемой скорости движения, а второй содержит сообщения с информацией о соседних агентах. Для тестирования алгоритма проведены эксперименты с тремя различными сценариями: со статическими препятствиями, смешанный, с динамическими агентами. Число эпизодов для обучения алгоритма при пяти агентах достигало 1500000. Моделирование движения робота на двух гусеницах в среде Gazebo показало, что в условиях статических препятствий робот достигает цели за наименьшее время. В присутствии динамических препятствий время увеличивалось в два раза по причине уклонения от столкновений. При этом расстояние до ближайшего агента оставалось безопасным (более 2 м). </p></abstract><trans-abstract xml:lang="en"><p>The work is devoted to the problem of controlling the movement of robots in a social environment in crowded places. An algorithm for planning the movement of mobile robots among stationary and moving obstacles using reinforcement learning has been developed and studied. The GA3C-CADRL algorithm was chosen as a prototype, in which the robot and obstacles are considered as interacting agents. The algorithm was modified and implemented using an LSTM recurrent neural network to approximate the value and policy functions simultaneously. The neural network was trained on a common data set obtained through actor-critic reinforcement learning. Additionally, the rl_ planner and social_msgs components have been developed to integrate a pre-trained planning algorithm into a robot control system on the Robot Operating System 2 software platform. The first component implements processing of input data, calculating the robot’s actions and generating the required speed of movement, and the second contains messages with information about neighboring agents. To test the algorithm, experiments were carried out with three different scenarios: – with static obstacles, – mixed, – with dynamic agents. The number of episodes for training the algorithm with 5 agents reached 1,500,000. Simulation of the movement of a robot on two tracks in the environment Gazebo showed that in conditions of static obstacles the robot reaches the goal in the shortest time. In the presence of dynamic obstacles, the time increased by 2 times due to collision avoidance. At the same time, the distance to the nearest agent remained safe (more than 2 meters).</p></trans-abstract><kwd-group xml:lang="ru"><kwd>мобильный робот</kwd><kwd>социальная среда</kwd><kwd>планирование движений</kwd><kwd>обучение с подкреплением</kwd><kwd>рекуррентная нейронная сеть</kwd></kwd-group><kwd-group xml:lang="en"><kwd>mobile robot</kwd><kwd>social environment</kwd><kwd>movement planning</kwd><kwd>reinforcement learning</kwd><kwd>recurrent neural network</kwd></kwd-group><funding-group><funding-statement xml:lang="ru">Исследование выполнено за счет гранта Российского научного фонда № 23-21-00287, https://rscf.ru/project/23-21-00287.</funding-statement><funding-statement xml:lang="en">This work was supported by the Grant of Russian Science Foundation #23-21-00287, https://rscf.ru/project/23-21-00287.</funding-statement></funding-group></article-meta></front><back><ref-list><title>References</title><ref id="cit1"><label>1</label><citation-alternatives><mixed-citation xml:lang="ru">Xiao X., Liu B., Warnell G., Stone P. Motion planning and control for mobile robot navigation using machine learning: a survey // Autonomous Robots. 2022. N. 46. P. 569—597. https://doi.org/10.48550/arXiv.2011.13112.</mixed-citation><mixed-citation xml:lang="en">Xiao X., Liu B., Warnell G., Stone P. Motion planning and control for mobile robot navigation using machine learning: a survey, Autonomous Robots, 2022, no. 46, pp. 569—597, available at: https://doi.org/10.48550/arXiv.2011.13112.</mixed-citation></citation-alternatives></ref><ref id="cit2"><label>2</label><citation-alternatives><mixed-citation xml:lang="ru">Koenig S., Likhachev M. D* Lite // Eighteenth National Conference on Artificial Intelligence. 2002. P. 476-483. URL: www.aaai.org.</mixed-citation><mixed-citation xml:lang="en">Koenig S., Likhachev M. D* Lite, Eighteenth National Conference on Artificial Intelligence, 2002, pp. 476—483, available at: www.aaai.org.</mixed-citation></citation-alternatives></ref><ref id="cit3"><label>3</label><citation-alternatives><mixed-citation xml:lang="ru">Филимонов А. Б., Филимонов Н. Б. Вопросы управления движением мобильных роботов методом потенциального наведения // Мехатроника, автоматизация, управление. 2019. Т. 20, № 11. С. 677—685.</mixed-citation><mixed-citation xml:lang="en">Filimonov A. B., Filimonov N. B. Issues of Motion Control of Mobile Robots Based on the Potential Guidance Method, Mekhatro nika, Avtomatizatsiya, Upravlenie, 2019, vol. 20, no. 11, pp. 677—685 (in Russian).</mixed-citation></citation-alternatives></ref><ref id="cit4"><label>4</label><citation-alternatives><mixed-citation xml:lang="ru">Fox D., Burgard W., Thrun S. The dynamic window approach to collision avoidance // IEEE Robotics &amp; Automation Magazine. 1997. N. 1 (4). P. 23—33.</mixed-citation><mixed-citation xml:lang="en">Fox D., Burgard W., Thrun S. The dynamic window approach to collision avoidance, IEEE Robotics &amp; Automation Magazine,1997, no. 1 (4), pp. 23—33.</mixed-citation></citation-alternatives></ref><ref id="cit5"><label>5</label><citation-alternatives><mixed-citation xml:lang="ru">Герасимов В. Н. Система управления движением мобильного робота в среде с динамическими препятствиями // Научно-технические ведомости СПбГПУ. Информатика, телекоммуникации и управление. СПб. 2013. № 5(181). С. 94—102.</mixed-citation><mixed-citation xml:lang="en">Gerasimov V. N. The motion control system of the mobile robot in environment with dynamic obstacles, Computing, Telecommunications and Control, 2013, no. 5 (181), pp. 94—102 (In Russian).</mixed-citation></citation-alternatives></ref><ref id="cit6"><label>6</label><citation-alternatives><mixed-citation xml:lang="ru">Macenski S., Martín F., White R., C lavero J. G. The Marathon 2: A Navigation System // 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS). 2020. P. 2718—2725. URL: https://doi.org/10.48550/arXiv.2003.00368.</mixed-citation><mixed-citation xml:lang="en">Macenski S., Martín F., White R., Clavero J. G. The Marathon 2: A Navigation System, 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2020, pp. 2718—2725, available at: https://doi.org/10.48550/arXiv.2003.00368.</mixed-citation></citation-alternatives></ref><ref id="cit7"><label>7</label><citation-alternatives><mixed-citation xml:lang="ru">Яндекс.Ровер на улицах Сколково [Электронный ресурс]. ООО "Яндекс". URL: https://yandex.ru/blog/company/yandeks-rover-na-ulitsakh-skolkovo (08.06.2023).</mixed-citation><mixed-citation xml:lang="en">Yandex.Rover on street of the Skolkovo [Electronic resource]. LTD "Yandex", available at: https://yandex.ru/blog/company/yandeks-roverna-ulitsakh-skolkovo (08.06.2023) (in Russian).</mixed-citation></citation-alternatives></ref><ref id="cit8"><label>8</label><citation-alternatives><mixed-citation xml:lang="ru">Alonso-Mora J., Andreas B., Martin R., Paul B., Roland S. Optimal Reciprocal Collision Avoidance for Multiple NonHolonomic Robots // Distributed Autonomous Robotic Systems: The 10th International Symposium. 2013. P. 203—216. URL: https://doi.org/10.1007/978-3-642-32723-0_15.</mixed-citation><mixed-citation xml:lang="en">Alonso-Mora J., Andreas B., Martin R., Paul B., Roland S. Optimal Reciprocal Collision Avoidance for Multiple NonHolonomic Robots, Distributed Autonomous Robotic Systems: The 10th International Symposium, 2013, pp. 203—216, available at: https://doi.org/10.1007/978-3-642-32723-0_15.</mixed-citation></citation-alternatives></ref><ref id="cit9"><label>9</label><citation-alternatives><mixed-citation xml:lang="ru">Ferrer G., Garrell A., Sanfeliu A. Layered costmaps for context-sensitive navigation // 2013 European Conference on Mobile Robots. 2013. P. 331—336. DOI: 10.1109/IROS20146942636.</mixed-citation><mixed-citation xml:lang="en">Ferrer G., Garrel A., Sanfeliu A. Layered costmaps for context-sensitive navigation, 2013 European Conference on Mobile Robots, 2013, pp. 331—336, available at: doi:10.1109/IROS20146942636.</mixed-citation></citation-alternatives></ref><ref id="cit10"><label>10</label><citation-alternatives><mixed-citation xml:lang="ru">Trautman P., Ma1 J., Richard M. Murray R. M., Krause A. Robot navigation in dense human crowds: the case for cooperation // 2013 IEEE International Conference on Robotics and Automation (ICRA). 2013. P. 2153—2160. URL: http://www.cds.caltech.edu/~murray/papers/tmmk13-icra.html.</mixed-citation><mixed-citation xml:lang="en">Trautman P., Ma1 J., Richard M. Murray R. M., Krause A. Robot navigation in dense human crowds: the case for cooperation, 2013 IEEE International Conference on Robotics and Automation, 2013, pp. 2153—2160, available at: http://www.cds.caltech.edu/~murray/papers/tmmk13-icra.html.</mixed-citation></citation-alternatives></ref><ref id="cit11"><label>11</label><citation-alternatives><mixed-citation xml:lang="ru">Rudenko A., Kucner T. P., Swaminathan C. S., Chadalavada R. T., Arras K. O., Lilienthal F. J. THÖR: Human-Robot Navigation Data Collection and Accurate Motion Trajectories Dataset // IEEE Robotics and Automation Letters. 2020. N. 2 (5). P. 676—682. URL: https://doi.org/10.48550/arXiv.1909.04403.</mixed-citation><mixed-citation xml:lang="en">Rudenko A., Kucner T. P., Swaminathan C. S., Chadalavada R. T., Arras K. O., Lilienthal F. J. THÖR: Human-Robot Navigation Data Collection and Accurate Motion Trajectories Dataset, IEEE Robotics and Automation Letters, 2020, no. 2 (5), pp. 676—682, available at: https://doi.org/10.48550/arXiv.1909.04403.</mixed-citation></citation-alternatives></ref><ref id="cit12"><label>12</label><citation-alternatives><mixed-citation xml:lang="ru">Tai L., Zhang J., Liu M., Burgard W. Socially Compliant Navigation Through Raw Depth Inputs with Generative Adversarial Imitation Learning // 2018 IEEE International Conference on Robotics and Automation (ICRA). 2018. P. 1111—1117. URL: https://doi.org/10.48550/arXiv.1710.02543.</mixed-citation><mixed-citation xml:lang="en">Tai L., Zhang J., Liu M., Burgard W. Socially Compliant Navigation Through Raw Depth Inputs with Generative Adversarial Imitation Learning, 2018, pp. 1111—1117, available at: https://doi.org/10.48550/arXiv.1710.02543.</mixed-citation></citation-alternatives></ref><ref id="cit13"><label>13</label><citation-alternatives><mixed-citation xml:lang="ru">Pérez-Higueras N., Caballero F., Merino L. Teaching Robot Navigation Behaviors to Optimal RRT Planners // International Journal of Social Robotics. 2018. N. 10. P. 235—249.</mixed-citation><mixed-citation xml:lang="en">Pérez-Higueras N., Cabalero F., Merino L. Teaching Robot Navigation Behaviors to Optimal RRT Planners, International Journal of Social Robotics, 2018, no. 10, pp. 235—249.</mixed-citation></citation-alternatives></ref><ref id="cit14"><label>14</label><citation-alternatives><mixed-citation xml:lang="ru">Plaat A. Deep Reinforcement Learning/ First Edition. Singapore: Springer Singapore, 2022. P. 406.</mixed-citation><mixed-citation xml:lang="en">Plaat A. Deep Reinforcement Learning, First Edition, Singapore, Springer Singapore, 2022, p. 406.</mixed-citation></citation-alternatives></ref><ref id="cit15"><label>15</label><citation-alternatives><mixed-citation xml:lang="ru">Long P., Fan T., Liao X., Liu W., Zhang H., Pan J. Towards Optimally Decentralized Multi-Robot Collision Avoidance via Deep Reinforcement Learning // 2018 IEEE International Conference on Robotics and Automation (ICRA). 2018. P. 6252—6259. URL: https://doi.org/10.48550/arXiv.1709.10082.</mixed-citation><mixed-citation xml:lang="en">Long P., Fan T., Liao X., Liu W., Zhang H., Pan J. Towards Optimally Decentralized Multi-Robot Collision Avoidance via Deep Reinforcement Learning, 2018 IEEE International Conference on Robotics and Automation (ICRA), 2018, pp. 6252—6259, available at: https://doi.org/10.48550/arXiv.1709.10082.</mixed-citation></citation-alternatives></ref><ref id="cit16"><label>16</label><citation-alternatives><mixed-citation xml:lang="ru">Everett M., Chen Y. F., How J. P. Collision Avoidance in Pedestrian-Rich Environments with Deep Reinforcement Learning // IEEE Access. 2021. P. 10357—10377. URL: https://doi.org/10.1109/ACCESS.2021.3050338.</mixed-citation><mixed-citation xml:lang="en">Everett M., Fan Chen Y., How J. P. Collision Avoidance in Pedestrian-Rich Environments with Deep Reinforcement Learning, IEEE Access, 2021, pp. 10357—10377, available at: https://doi.org/10.1109/ACCESS.2021.3050338.</mixed-citation></citation-alternatives></ref><ref id="cit17"><label>17</label><citation-alternatives><mixed-citation xml:lang="ru">Chen Y. F., Liu M., Everett M., How J. P. Decentralized non-communicating multiagent collision avoidance with deep reinforcement learning // 2017 IEEE International Conference on Robotics and Automation (ICRA). 2017. P. 285—292. URL: https://doi.org/10.48550/arXiv.1609.07845.</mixed-citation><mixed-citation xml:lang="en">Chen Y., Liu M., Everett M., How J. P. Decentralized non-communicating multiagent collision avoidance with deep reinforcement learning, 2017 IEEE International Conference on Robotics and Automation (ICRA), 2017, pp. 285—292, available at: https://doi.org/10.48550/arXiv.1609.07845.</mixed-citation></citation-alternatives></ref><ref id="cit18"><label>18</label><citation-alternatives><mixed-citation xml:lang="ru">Zhou Z., Zhu P., Zeng Z., Xiao J., Lu H. Robot navigation in a crowd by integrating deep reinforcement learning and online planning // Applied Intelligence. 2022. P. 15600—15616. URL: https://doi.org/10.48550/arXiv.2102.13265.</mixed-citation><mixed-citation xml:lang="en">Zhou Z., Zhu P., Zeng Z., Xiao J., Lu H. Robot navigation in a crowd by integrating deep reinforcement learning and online planning, Applied Intelligence, 2022, pp. 15600—15616, available at: https://doi.org/10.48550/arXiv.2102.13265.</mixed-citation></citation-alternatives></ref><ref id="cit19"><label>19</label><citation-alternatives><mixed-citation xml:lang="ru">Macenski S., Foote T., Gerkey B., Lalancette C., Woodall W. Robot Operating System 2: Design, architecture, and uses in the wild // Science Robotics. 2022. N. 7 (66). URL: https://doi.org/10.1126/scirobotics.abm6074.</mixed-citation><mixed-citation xml:lang="en">Macenski S., Foote T., Gerkey B., Lalancette C., Woodall W. Robot Operating System 2: Design, architecture, and uses in the wild, Science Robotics, 2022, no. 7 (66), available at: https://doi.org/10.1126/scirobotics.abm6074.</mixed-citation></citation-alternatives></ref><ref id="cit20"><label>20</label><citation-alternatives><mixed-citation xml:lang="ru">Ларионов А. А. Планирование движения мобильного робота в социальной среде с обучением с подкреплением. Магистерская диссертация, СПбПУ [Электронный ресурс]. URL: https://github.com/TonyCooT/msc_thesis (08.06.2023).</mixed-citation><mixed-citation xml:lang="en">Larionov A. A. Master thesis "Planning mobile robot movement in social environment with reinforcement learning" [Electronic resource], LTD "Yandex", available at: https://github.com/TonyCooT/msc_thesis (08.06.2023) (In Russian).</mixed-citation></citation-alternatives></ref></ref-list><fn-group><fn fn-type="conflict"><p>The authors declare that there are no conflicts of interest present.</p></fn></fn-group></back></article>
