基于模糊滑模的多机器人编队控制研究

Formation control is a basic problem in the multi-robots systems. In this project, fast and accurate formation control algorithms are designed by the combination of sliding mode control methods and fuzzy logic theory. Firstly, considering the uncertainties of the complex environment and the dynamics of the motors, a fuzzy dynamic model of the tracking error is established, based on which a fuzzy sliding surface is designed and a control law is proposed to guarantee the tracking error to converge to zero. Secondly, a singular system model is formulated for multi robots system by considering the distance or angle constraints of the formation. Then, the sliding mode control problem of singular system is studied and a criterion is proposed to guarantee the system to be regular, impulse free and stable. Thirdly, the problem of communication constraints is investigated. A stochastic model of the communication channel is established. By considering the data loss rate and transmission delays, a new sliding surface is designed and a sliding observer is given to estimate the system state, based on which, the sliding controller law with high level performance is proposed. The relationships of the system performance and the data loss rate or transmission delays are obtained. Finally, the methods and main results of this project are verified by both simulations and experiments.

编队控制是多机器人系统中广泛存在的一类基础性问题。本项目拟将滑模控制与模糊控制相结合，实现快速、高精度的多机器人编队控制。首先，考虑系统的不确定因素和驱动电机的动态特性，建立跟踪误差的模糊模型，设计具有较高精度的滑模控制算法，给出保证跟踪误差收敛的条件；其次，考虑编队队形约束，建立多机器人系统的奇异系统模型，并基于奇异滑模控制理论设计编队控制算法，给出保证系统正则、无脉冲、稳定的条件；然后，针对通讯受限问题，建立通讯信道的传输模型，设计基于通讯信道特性的模糊滑模面，给出滑模观测器的设计方法，并在此基础上设计具有较高精度的编队控制算法，建立系统性能与网络丢包率、时延的关系。最后，利用Matlab仿真软件和多机器人实验平台进行算法验证。

轮式移动机器人具有机械结构简单、灵活度大、操作性好等优点，在社会生产、生活的各个方面得到了广泛的应用。随着机器人系统需要解决的问题与任务也越来越复杂，实时性要求也越来越高。单靠增加单个机器人的功能，会导致系统变得复杂庞大，成本较高。多机器人系统通过机器人之间的相互协作，可以有效的解决这类难题，同时，提升了系统的鲁棒性和容错性。. 本项目针对机器人系统的轨迹跟踪与编队问题进行研究，提出了一种基于航迹推测的视觉定位方法，提高了视觉定位速度，降低了硬件成本；给出了轮式移动机器人运动学模型参数不确定和车轮纵向打滑情况下的自适应轨迹跟踪控制算法，提高了轨迹跟踪精度；设计了基于领航-跟随法的双幂次滑模编队控制算法，实现了多机器人的快速编队；提出了基于事件触发机制的静态输出反馈控制器设计方法，有效降低了网络通讯负载。本项目发表论文 15 篇，其中SCI/EI 论文 11 篇；申请发明专利 6项，其中授权3项；申请软件著作权1项。本项目辅助培养博士生1人，辅助培养硕士生3人；培养硕士研究生9人，其中毕业6人，在读3人。