多机械臂煤矸石分拣机器人是实现煤矸石分拣智能化的重要发展方向。煤矸石质量大小不一、数量多、分布随机，传统的机械臂定点分拣方式和基于分拣时间最短的分拣策略虽然简单易行，但机械臂返回初始点消耗时间过多，分拣策略也不能保证系统综合效益。简单的机械臂线性叠加会占用更大工作空间，导致煤矸石的分拣效率低下。为了提高多机械臂煤矸石分拣机器人的分拣效率，本文在国家自然科学基金面上项目《煤矿井下多机械臂智能协同分拣煤矸石基础研究》项目的资助下，提出一种多机械臂动态工作空间协同分拣方法。

针对煤矸石运动速度快、质量大，传统定点分拣方式的分拣路径对分拣效率影响大，任务最优分配方法对分拣时间、稳定性等要求高的问题，深入分析机械臂的分拣过程，构建了基于机械臂动态目标跟踪的分拣模型，提出三环PID轨迹跟踪算法和动态放置方式分拣方法，通过求解动态目标抓取点和分拣时间等参数，实现了机械臂快速跟踪并完成分拣工作。实验结果表明，该方法能够满足冲击小、时间优等要求，分拣时间较传统定点放置缩短约0.2s。

针对任务分配策略不佳导致分拣率低的问题，对比基于单一因素的任务分配策略，从机械臂分拣矸石的收益和代价两方面考虑，建立了综合效益函数，构建了基于时间、质量最优的多目标任务分配模型。实验结果表明，基于综合效益最大的任务分配策略较基于时间最短和基于质量最大的任务分配策略分拣率高8%。

针对煤矸石分布随机、任务量大和多机械臂协同效率低等问题，提出了基于效益函数的多机械臂动态工作空间动态放置点的分拣机制。构建了基于动态工作空间多机械臂协同分拣模型，先通过改进匈牙利算法获得任务分配优先级，保证系统效益，再通过任务协同与轨迹协同消除多机械臂之间的冲突，实现多机械臂高效协同分拣煤矸石。实验结果表明，所提方法能充分发挥机械臂的分拣性能，提高多机械臂协同能力，比传统基于时间最短的定点分拣方式的分拣率高20%。

Multi-arm coal gangue sorting robot is an important development direction to realize intelligent coal gangue sorting. Coal gangue has different quality sizes, large quantities, and random distribution. Although the traditional fixed-point sorting method of the mechanical arm and the sorting strategy based on the shortest sorting time is simple and feasible, the mechanical arm's return to the initial point consumes too much time, and the sorting strategy cannot guarantee the comprehensive benefits of the system. Simple linear superposition of mechanical arms will occupy a larger working space, resulting in low sorting efficiency of coal gangue. To improve the sorting efficiency of a multi-arm coal gangue sorting robot, a multi-arm dynamic workspace collaborative sorting method was proposed in this paper under the support of the National Natural Science Foundation project "Basic Research on Intelligent Cooperative Sorting of Coal Gangue by Multi-arm in Coal Mine".

Aiming at the problems that coal gangue moves fast and has high quality, the sorting path of traditional fixed-point sorting has great influence on sorting efficiency, and the task optimal allocation method has high requirements on sorting time and stability, the sorting process of mechanical arm is deeply analyzed, a sorting model based on the dynamic target tracking of mechanical arm is constructed, and the three-ring PID trajectory tracking algorithm and dynamic placement sorting method are proposed. By solving the parameters of dynamic target grasping point and sorting time, the robot arm fast tracking and sorting were realized. The experimental results show that this method can meet the requirements of low impact and good time, and the sorting time is shortened by about 0.2s compared with the traditional fixed-point placement.

Aiming at the problem of low sorting rate caused by poor task allocation strategy, compared with task allocation strategy based on a single factor, a comprehensive benefit function was established from the perspective of income and cost of gangue sorting by mechanical arm, and a multi-objective task allocation model based on optimal time and quality was constructed. The experimental results show that the sorting rate of the task assignment strategy based on the maximum comprehensive benefit is 8% higher than that based on the shortest time and the maximum quality.

Aiming at the problems of random distribution of coal gangue, a large number of tasks, and low collaborative efficiency of multi-manipulator, a sorting mechanism of dynamic placement points of multi-manipulator dynamic workspace based on benefit function was proposed. A multi-arm collaborative sorting model based on a dynamic workspace was constructed. Firstly, the Hungarian algorithm was improved to obtain the priority of task allocation and ensure the system's benefit. Then, the conflict between the multi-arm was eliminated through task collaboration and trajectory collaboration, and the multi-arm collaborative sorting of coal gangue was realized efficiently. The experimental results show that the proposed method can give full play to the sorting performance of the robotic arm and improve the cooperative ability of multiple robotic arms. The sorting rate is 20% higher than that of the traditional fixed-point sorting method based on the shortest time.

[1]陈雪梅,张晞,徐莉莉,等.煤与矸石分形维数的差异研究[J].煤炭科学技术,2017,45(07):196-199.

[2]Wang Z, Xie S, Chen G, et al. An Online Flexible Sorting Model for Coal and Gangue Based on Multi-Information Fusion[J]. IEEE Access, 2021, 9: 90816-90827.

[3]周楠,姚依南,宋卫剑,等.煤矿矸石处理技术现状与展望[J].采矿与安全工程学报,2020,37(01):136-146.

[4]卫玉花.“双碳”目标背景下我国选煤行业发展现状与展望[J].煤炭加工与综合利用,2022(03):13-15.

[5]郭静,李磊,李志明.干法选煤技术创新进展及其节能节水降污效果分析[J].中国煤炭,2022,48(05):68-75.

[6]庞义辉,王国法,仁怀伟.智慧煤矿主体架构设计与系统平台建设关键技术[J].煤炭科学技术,2019,47(03):35-42.

[7]朱雨贺. 基于深度学习目标检测的机械臂分拣系统研发[D].浙江大学,2019.

[8]刘璇,吴兆敏,王唱,等.酒醅抓取机械臂轨迹规划研究[J].机械传动,2022,46(09):28-35+48.

[9]张袁浩,潘祥生,陈晓晶,等.智能选矸机器人关键技术研究[J].工矿自动化,2022,48(06):69-76+111.

[10]Watt J S, Steffner E J. Dual energy gamma-ray transmission techniques applied to on-line analysis in the coal and mineral industries[J]. The International Journal of Applied Radiation and Isotopes, 1985, 36(11): 867-877.

[11]Hobson D M, Carter R M, Yan Y, et al. Differentiation between coal and stone through image analysis of texture features[C]//2007 IEEE International Workshop on Imaging Systems and Techniques. IEEE, 2007: 1-4.

[12]Singh V, Singh T N, Singh V. Image processing applications for customized mining and ore classification[J]. Arabian Journal of Geosciences, 2011, 4: 1163-1171.

[13]Dou D, Wu W, Yang J, et al. Classification of coal and gangue under multiple surface conditions via machine vision and relief-SVM[J]. Powder Technology, 2019, 356: 1024-1028.

[14]Sun Z, Li D, Huang L, et al. Construction of intelligent visual coal and gangue separation system based on CoppeliaSim[C]//2020 5th International Conference on Automation, Control and Robotics Engineering (CACRE). IEEE, 2020: 560-564.

[15]Li R, Wang R, Tian X. Binocular vision mechanical arm system based on salient region target recognition[C]//Journal of Physics: Conference Series. IOP Publishing, 2020, 1550(3): 032059.

[16]Shang D, Wang Y, Yang Z, et al. Study on comprehensive calibration and image sieving for coal-gangue separation parallel robot[J]. Applied Sciences, 2020, 10(20): 7059.

[17]韩成石,董长双,周西军,等.煤和矸石 γ 射线分选系统的研究[J].山西矿业学院学报,1997(2):157-161.

[18]孔力,李红.基于双能 γ 射线的煤,矸石区域分割识别方法[J].华中科技大学学报(自然科学版),1998(1):39-40.

[19]孔力,李红,徐恕宏,等.双能γ射线透射法煤矸石在线识别与分选系统[J].华中理工大学学报,1997(10):108-109+113.

[20]乔力江汉,何克焓.基于分形维数及细观孔隙结构特征的煤矸石识别研究[J].中国矿业,2021,30(09):120-125.

[21]王鹏,曹现刚,夏晶,等.基于机器视觉的多机械臂煤矸石分拣机器人系统研究[J].工矿自动化,2019,45(09):47-53.

[22]郜亚松,张步勤,郎利影.基于深度学习的煤矸石识别技术与实现[J].煤炭科学技术,2021,49(12):202-208.

[23]梁兴国,李云峰,李燕,等.智能干选技术研究应用及发展趋势[J].选煤技术,2019(01):92-96+102.

[24]缪杰.煤矸石智能分拣机器人的研究与应用[J].智能矿山,2023,4(01):58-62.

[25]亢健东.煤矸石分拣 机器人控制系统研究[J].中国石油和化工标准与质量,2021,41(06):155-157.

[26]李浩,文彪,左学海等.煤矸石分拣机械手的系统设计及实验模拟[J].煤,2023,32(04):21-23.

[27]赵明辉,宣鹏程,张少宾.并联煤矸石分拣机器人的结构设计及分析[J].机床与液压,2021,49(05):55-59.

[28]曹现刚,费佳浩,王鹏,等.基于多机械臂协同的煤矸分拣方法研究[J].煤炭科学技术,2019,47(04):7-12.

[29]张永超,于智伟,丁丽林.基于强化学习的煤矸石分拣机械臂智能控制算法研究[J].工矿自动化,2021,47(01):36-42.

[30]刘鹏,马宏伟,乔心州等.柔索驱动拣矸机器人最小索拉力等值曲面研究[J].西安科技大学学报,2020,40(05):797-804.

[31]乔心州,武琛琛,刘鹏,等.基于故障树的柔索驱动拣矸机器人系统分拣可靠性研究[J].工矿自动化,2022,48(08):107-113..

[32]何泽强.基于机器视觉的工业机器人分拣系统设计[D].哈尔滨工业大学,2016.

[33]王易凤.机器人动态目标抓取的跟踪控制研究[D].南京理工大学,2019.

[34]Gonçalves A R C, Guerrero-Pena E, Araújo A F R, et al. Evolutionary Trajectory Planning with Obstacles for a Mobile Manipulator[J]. IFAC-PapersOnLine, 2020, 53(2): 7921-7926.

[35]Shin K,McKay N. Minimum-time Control of Robotic Manipulators with Geometric Path Constraints[J].IEEETransactions on Automatic Control, 1985,30(6):531-541.

[36] Bobrow J E, Dubowsky S,Gibson J.Time-optimal Control of robotic manipulators along specified paths[J]. The international journal of robotics research, 1985,4(3):3-17.

[37]肖文皓,白瑞林,许凡,等.基于信赖域算法的机械臂时间最优轨迹规划[J].传感器与微系统,2013,32(06):77-80.

[38]赵铁军,王玲.基于改进贪心算法的Delta机器人分拣路径优化[J].组合机床与自动化加工技术,2021(12):58-61+66.

[39]毕宪东,王振,李朝龙.基于Delta机器人的食品生产线动态目标抓取方法[J].食品与机械,2022,38(06):117-122.

[40]乔心州,龚莉,刘鹏.柔索驱动拣矸机器人抓取轨迹规划研究[J].机械科学与技术:1-10[2022-11-15].

[41]张俊,刘浩阳,许涛.Delta机器人运动轨迹动态规划方法[J].机械设计与研究,2022,38(03):81-87.

[42]苏婷婷,张好剑,王云宽,等.基于费拉里法的Delta机器人动态目标抓取算法[J].华中科技大学学报(自然科学版),2018,46(06):128-132.

[43]Seddaoui A, Saaj C M. Collision-free optimal trajectory generation for a space robot using genetic algorithm[J]. Acta Astronautica, 2021, 179: 311-321.

[44]Huang T, Wang P F, Mei J P, et al. Time minimum trajectory planning of a 2-DOF translational parallel robot for pick-and-place operations[J]. CIRP annals, 2007, 56(1): 365-368.

[45]李振娜,王涛,王斌锐,等.基于带约束S型速度曲线的机械手笛卡尔空间轨迹规划[J]．智能系统学报,2019,14( 4) : 655-661.

[46]Rubio F J, Valero F J, Suñer J L, et al. Simultaneous algorithm to solve the trajectory planning problem[J]. Mechanism and Machine Theory, 2009, 44(10): 1910-1922.

[47]Shirzadeh M, Amirkhani A, Tork N, et al. Trajectory tracking of a quadrotor using a robust adaptive type-2 fuzzy neural controller optimized by cuckoo algorithm[J]. ISA transactions, 2021, 114: 171-190.

[48]倪鹤鹏,刘亚男,张承瑞,等.基于机器视觉的Delta机器人分拣系统算法[J].机器人,2016,38(01):49-55.

[49]丁鑫,张锁怀,祝梦洁,等.模糊PID控制的桁架机器人轨迹规划研究[J].机床与液压,2020,48(03):9-13.

[50]蔡倩倩.基于机器视觉的馒头分拣机器人轨迹规划研究[D].河南工业大学,2019.

[51]李卓隆,任永杰,林嘉睿.基于在线预测和规划的机器人动态跟踪抓取方法[J].自动化与仪器仪表,2021(10):183-189.

[52]夏晶,张昊,周世宁,等.煤矸石分拣 机器人动态拣取避障路径规划[J].煤炭学报,2021,46(S1):570-577.

[53]CROFT E. A,FENTON R. G,BENHABIB B.Optimal rendezvous-point selection for robotic interception of moving objects[J].IEEE Tranactions on Systems, Man,and Cybernetics,1998,28(2):192-204.

[54]张朝阳,周惠兴,曹荣敏,等.基于金字塔形寻优的传送带动态抓取研究[J].制造业自动化,2015,37(9):92-95.

[55]ZHANG Zhengyou.Camera calibration with one-dimensional objects[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2004,26(7):892-899.

[56]吴笑萍.一类多机器人系统任务分配方法的研究[D].南京理工大学,2012.

[57]袁晗,童上高,顾文昊,等.机器人动态物料追随抓取技术的轨迹算法设计[J].机器人技术与应用,2019(01):29-32.

[58]唐苏妍,朱一凡,李群,等.多Agent系统任务分配方法综述[J].系统工程与电子技术,2010,32(10):2155-2161.

[59]Nunes E, Manner M, Mitiche H, et al. A taxonomy for task allocation problems with temporal and ordering constraints[J]. Robotics and Autonomous Systems, 2017, 90: 55-70.

[60]Gerkey B P, Matarić M J. A formal analysis and taxonomy of task allocation in multi-robot systems[J]. The International journal of robotics research, 2004, 23(9): 939-95.

[61]Werger B B, Mataric M J. Broadcast of local eligibility: behavior-based control for strongly cooperative robot teams[C]//Proceedings of the fourth international conference on autonomous agents. 2000: 21-22.

[62]Owen W S, Croft E A, Benhabib B. A multi-arm robotic system for optimal sculpting[J]. Robotics and Computer-Integrated Manufacturing, 2008, 24(1): 92-104.

[63]Smith R G. The contract net protocol: High-level communication and control in a distributed problem solver[J]. IEEE Transactions on computers, 1980, 29(12): 1104-1113.

[64]宁宇铭,李团结,姚聪等.基于快速扩展随机树―贪婪边界搜索的多机器人协同空间探索方法[J].机器人,2022,44(06):708-719.

[65]张进,郭浩,陈统.基于可适应匈牙利算法的武器-目标分配问题[J].兵工学报,2021,42(06):1339-1344.

[66]刘勇,陈国东,周游.基于单纯形法的LINGO求解一般指派问题的探讨[J].中国管理信息化,2008(06):86-87.

[67]Atay N,Bayazit B.Mixed-integer linear programming solution to multi-robot task allocation problem[J].Washington Univ,2006．

[68]Zhu H,Zhou M.C and R.Alkins(2012).Group Role Assignment via a Kuhn-Munkres Algorithm-based Solution. IEEE Transactions on Systems,Man and Cybemetics,Part A:Systems and Humans,42,(3):739-750.

[69]Colorni A, Dorigo M, and Maniezzo V Distributed optimization by ant colonies[C]. The First European Conference on Artificial Life, Elsevier, 1992:134-142.

[70]曹宗华,吴斌,黄玉清,等.基于改进蚁群算法的多机器人任务分配[J].组合机床与自动化加工技术,2013,No.468(02):34-37.

[71]姜健,臧希喆,闫继宏,等.基于一种蚁群算法的多机器人动态感知任务分配[J].机器人,2008(03):254-258+263.

[72]Hou E S H, Ansari N, Ren H. A genetic algorithm for multiprocessor scheduling[J]. IEEE Transactions on Parallel and Distributed systems, 1994, 5(2): 113-120.

[73]蒋莹,肖晓明,蔡自兴,等.基于改进蚁群算法的多机器人任务分配方法[J].计算机测量与控制,2008(06):827-829+842.

[74]Tan Y,zhu Y.Fireworks Algorithm for Optimization[C]//Beijing,China,2010lnternational Conference on Advancesin Swarm lntelligence.Springer-Verlag,2010:355-364.

[75]张涛,刘天威,李富章,等.基于改进烟花算法的多目标多机器人任务分配[J].信号处理,2020,36(08):1243-1252.

[76]郑轶.多智能机器人系统任务分配与协作研究[D].南京理工大学,2004.

[77]包翔宇,曹学鹏,张弓,等.多机器人协同系统的研究综述及发展趋势[J].制造技术与机床,2019(11):26-30.

[78]侯震.基于激光雷达的多机器人系统协同控制问题研究[D].吉林大学,2019.

[79]Cao Y U, Kahng A B, Fukunaga A S. Cooperative mobile robotics: Antecedents and directions[J]. Robot colonies, 1997: 7-27.

[80]秦新立.多机器人协同任务分配与路径规划的研究[D].天津大学,2018.

[81]付梦家,游晓明.多机器人系统及其路径规划方法综述[J].软件导刊,2017,16(01):177-179.

[82]周勇城.基于多智能体的多机械臂系统自适应协调控制及其应用[D].重庆大学,2016.

[83]Kube C R, Zhang H. Collective robotics: From social insects to robots[J]. Adaptive behavior, 1993, 2(2): 189-218.

[84]Xu Y, Mendenhall S, Ha V, et al. Herding nerds on your table: NerdHerder, a mobile augmented reality game[M]//CHI'12 Extended Abstracts on Human Factors in Computing Systems. 2012: 1351-1356.

[85]Towle Jr B A, Nicolescu M. An auction behavior-based robotic architecture for service robotics[J]. Intelligent Service Robotics, 2014, 7(3): 157-174.

[86]陈昕叶.动态环境下多机器人协同搜救任务分配方法研究[D].华南理工大学,2020.

[87]曹现刚,吴旭东,王鹏,等.面向煤矸石分拣 机器人的多机械臂协同策略[J].煤炭学报,2019,44(S2):763-774.

[88]曹现刚,乔欢乐,吴旭东,等.考虑含矸率时变性的多臂协同策略优化方法[J].机械科学与技术:1-9[2023-03-21].

[89]毛勇.控制电机转速的PID算法综述[J].机电信息,2011(06):43-44.

[90]安泽,熊芬芬,梁卓楠.基于偏置比例导引与凸优化的火箭垂直着陆制导[J].航空学报,2020,41(05):247-260.