面向抛掷分拣的工业机器人系统设计与运动轨迹规划算法

       面对日益增长的城市生活垃圾，世界多个国家和地区提出了垃圾资源化处理的发展要求。近十年来，随着机器人技术、视觉处理技术、人工智能等迅速发展，机器人分拣系统在一众垃圾自动化分拣系统中脱颖而出。相对于机器人在传统工业领域的大量应用， 其在垃圾分拣场景中的应用尚不成熟。
       抛体运动是一种生活中常见的经典力学现象， 针对垃圾分拣应用追求分拣效率、物品不怕磕碰的特点，本文提出了一种利用机器人抛掷物品来提高分拣效率的研究猜想。为了验证这一猜想，本文设计了一套面向生活垃圾分拣的工业机器人系统， 并详细介绍了该系统的设计过程、实现细节和性能表现。 着重探索机器人高速分拣的轨迹规划算法， 基于梯形配点法实现了时间最优的抛掷分拣轨迹优化算法，通过具体指标确认了该算法具备出色的时间寻优能力。
       为了对比分析抛掷分拣方式与传统抓取放置方式的差异，设计具体分拣任务，通过算法求解数据来对比两种方式的轨迹耗时，通过样机实验来观察两种方式的真实分拣表现。结果表明，抛掷分拣方式的轨迹耗时约占抓取放置方式同阶段耗时的 60%，具备更高的分拣效率；算法生成的抛掷分拣轨迹在样机实验中的分拣成功率高达 99%，能够实现生活中一大类物品的高速抛掷分拣。

       With the growing amount of municipal solid waste, many countries and regions have put forward development requirements for waste recycling. In the past ten years, with the rapid development of robotics, machine vision and artificial intelligence, robotic sorting system stands out among the automatic waste sorting systems. Compared with many robotics applications in traditional industrial fields, their application in waste sorting filed is not yet mature.
       Projection motion is a classical mechanical phenomenon in life. Given the characteristics of waste sorting applications that pursue sorting efficiency and that items are not afraid of breaking, this thesis proposes a research idea to improve sorting efficiency by using robots to toss items. In order to verify this idea, this paper designs an industrial robot system for waste sorting, and
introduces the design process, implementation details and performance in detail. Focusing on exploring the trajectory planning algorithm of robot high-speed sorting, a time-optimal toss sorting trajectory optimization algorithm based on the trapezoidal collocation method is implemented. It is confirmed that the algorithm has an excellent time-optimizing ability through specific indicators.
       In order to compare and analyze the difference between the toss sorting method and the traditional pick-and-place method, specific sorting tasks are designed, the trajectory time-consuming is compared by algorithm solutions, and the actual sorting performance is observed through real-machine experiments. The results show that the toss sorting method has higher sorting efficiency--its trajectory time-consuming is about 60% of that of the pick-andplace method; the toss sorting trajectory generated by the algorithms can achieve high-speed toss sorting of a large category of items in life with a sorting success rate of as high as 99%.

[1] GUNDUPALLI S P, HAIT S, THAKUR A. A review on automated sorting of sourceseparated municipal solid waste for recycling[J]. Waste Management,2017,60:56–74.
[2] 李磊,袁光钰.中国城市生活垃圾处理现状及展望[J].世界环境,2017,169(6):24-27.
[3] SARC R, CURTIS A, KANDLBAUER, et al. Digitalisation and intelligent robotics in value chain of circular economy oriented waste management–A review[J]. Waste Management,2019,95:476–492.
[4] 朱莺莺 .中国生活垃圾处理技术应用现状及未来主流技术探讨[J].台州学院学报,2017,39(3).
[5] ZenRotics. Solutions[OL]. https://zenrobotics.com/.
[6] OP Teknik. SELMA Robot Sorting[OL]. https://www.opteknik.se/sorteringssida?lang=en.
[7] AMP Robotics. Products[OL]. https://www.amprobotics.com/.
[8] BHS. Max-AI[OL]. https://www.max-ai.com/
[9] RAPTOPOULOS F, KOSKINOPOULOU M, MANIADAKIS M. Robotic Pick-andToss Facilitates Urban Waste Sorting[C]. IEEE International Conference on Automation Science and Engineering,2020,2020-Augus:1149–1154.
[10] ZENG A, SONG S, Lee J, et al. TossingBot: Learning to Throw Arbitrary Objects with Residual Physics[J]. IEEE Transactions on Robotics,2020,36(4):1307 –1319.
[11] KIM S, DONCIEUX S. Learning highly diverse robot throwing movements through quality diversity search[C]. GECCO 2017-Proc. Genet. Evol. Comput. Conf. Companion,2017,vol.July15-19:1177–1178.
[12] HU J S, CHIEN M C, CHANG Y J, et al. A ball-throwing robot with visual feedback[C]. IEEE/RSJ 2010 International Conference on Intelligent Robots and Systems, IROS 2010 - Conference Proceedings,2010:2511–2512.
[13] FRANK H, WELLERDICK-WOJTASIK N, HAGEBEUKER B, et al. Throwing objects - A bio-inspired approach for the transportation of parts[C]. 2006 IEEE International Conference on Robotics and Biomimetics, ROBIO 2006:91 –96.
[14] FRANK H, BARTEIT D, WELLERDICK-WOJTASIK N, et al. Autonomous mechanical controlled grippers for capturing flying objects[C]. IEEE International Conference on Industrial Informatics (INDIN),2007,1:431–436.
[15] FRANK H. Design and simulation of a numerical controlled throwing device[C]. Proc. - 2nd Asia Int. Conf. Model. Simulation, AMS 2008,2008:777–782.
[16] BARTEIT D, FRANK H, KUPZOG F. Accurate prediction of interception positions for catching thrown objects in production systems[C]. IEEE Int. Conf. Ind. Informatics,2008:893–898.
[17] FRANK H. Determination of launching parameters for throwing objects in logistic processes with direct hits[C]. IEEE Int. Conf. Emerg. Technol. Fact. Autom. ETFA,2008,2:58–61.
[18] FRANK H, BARTEIT D, MEYER M, et al. Optimized control methods for capturing flying objects with a Cartesian Robot[C]. 2008 IEEE Int. Conf. Robot. Autom. Mechatronics, RAM 2008,2008,00:160–165.
[19] FRANK H, BARTEIT D, KUPZOG F. Throwing or shooting - A new technology for logistic chains within production systems[C]. 2008 IEEE International Conference on Technologies for Practical Robot Applications, TePRA, 2008:62–67.
[20] FRANK H, MITTNACHT A, SCHEIERMANN J. Throwing of Cylinder-Shaped Objects[C]. 2009 IEEE/ASME International Conference on Advanced Intelligent Mechatronics,2009,1:59–64.
[21] FRANK H. Throwing of Objects: A New Technology for Factory Automation[C]. 2009 First International Conference on Computational Intelligence,2009:4–5.
[22] FRANK H, MITTNACHT A,MOSCHINSKY T, et al. 1-DOF-robot for fast and accurate throwing of objects[C]. ETFA 2009 - 2009 IEEE Conference on Emerging Technologies and Factory Automation, 2009.
[23] PONGRATZ M, KUPZOG F, FRANK H, et al. Transport by throwing - A bioinspired approach[C]. IEEE International Conference on Industrial Informatics (INDIN),2010:685–689.
[24] FRANK H, FRANK T, MITTNACHT A, et al. A bioinspired 2-DOF throwing robot[C]. IEEE AFRICON Conference,2011,September,:13–15.
[25] SENOO T, NAMIKI A, ISHIKAWA M. High-speed throwing motion based on kinetic chain approach[C]. 2008 IEEE/RSJ International Conference on Intelligent Robots and Systems,IROS,2008:3206–3211.
[26] KIM C H, SUGANO S. Executing optimized throwing motion on robot arm with free joint[J]. Adv. Robot.,2016,30(24)1571–1578.
[27] 张欣.智能乒乓球发球机器人[D].厦门:厦门大学,2014.
[28] 郝亚非.抛体运动轨迹的数值分析[J].物理通报,2015,12:10-12.
[29] 楚安夫.关于斜抛运动的分析[J].大学物理,1997,16(9):46-47.
[30] 吴昌林,张卫国,姜柳林.机械设计[M].第三版.武汉:华中科技大学出版社,2010:372-390.
[31] WILSON M.机器人系统实施-制造业中的机器人、自动化和系统集成[M].王伟,贠超,译.北京:机械工业出版社,2016:16-25.
[32] 王军锋,唐宏.伺服电机选型的原则和注意事项[J].装备制造技术,2009,11:129-133.
[33] 黄捷建,张静,李浩,等.伺服电机负载惯量比的合理取值[J].微电机,2017,50(11):72-75.
[34] 蔡一.交流伺服系统高速定位控制的研究[D].哈尔滨:哈尔滨工业大学,2009.
[35] 控 制 工 控 网 . 交 流 伺 服 驱 动 器 的 控 制 结 构 及 基 本 整 定 [OL]. http://article.cechina.cn/21/0518/08/20210518083621.htm.
[36] 赵刚.数控机床交流伺服系统动态性能分析与参数整定技术研究[D].天津:天津大学,2010.
[37] 张洪国.高速电机的高精度控制系统及其算法的研究 [D].哈尔滨:哈尔滨工业大学,2013.
[38] 李虎修.交流伺服系统高性能速度控制关键技术研究[D].山东:山东大学,2012.
[39] 肖维荣,齐 蓉.装备自动化-工程设计与实践[M].第 2 版.北京:机械工业出版社,2021:46-62.
[40] 彭瑜,何衍庆.运动控制系统软件原理及其标准功能块应用[M].北京:机械工业出版社,2019:2.
[41] 李瑞峰.时间最优的机器人点到点執迹规划问题研究[D].大连:大连理工大学,2016.
[42] KELLY M. An introduction to trajectory optimization: How to do your own direct collocation[J]. SIAM Rev.,2017,59(4):849–904.
[43] KELLY M P. Transcription Methods for Trajectory Optimization A beginners tutorial[J]. arXiv,2015:1–14.
[44] PAPADAKIS E, RAPTOPOULOS F, KOSKINOPOULOU M, et al. On the Use of Vacuum Technology for Applied Robotic Systems[C]. International Conference on Mechatronics and Robotics Engineering, ICMRE 2020, Feb. 2020:73–77.
[45] SMC(中 国)有 限公司. 现代实用气动技术[M]. 第3版.北京:机械工业出版社,2008:538-540.
[46] KOSKINOPOULOU M, RAPTOPOULOS F, PAPADOPOULOS G, et al. Robotic Waste Sorting Technology: Toward a Vision-Based Categorization System for the Industrial Robotic Separation of Recyclable Waste[J]. IEEE Robot. Autom. Mag., 2021,June:50-60.