基于生物智能算法的群体机器人协同控制

发布时间：2018-05-09 04:04

本文选题：群体机器人 + 协同控制　；参考：《东华大学》2016年博士论文

【摘要】：随着机器人的使用模式从部件式单元应用向系统式应用的方向发展,机器人的应用已经逐渐地进入了大众的视野,同时群体机器人系统因其能够更好地满足人们的需求而得到了越来越多的关注与研究,相关的研究与应用也如雨后春笋般层出不穷。本论文介绍了群体机器人技术的起源、发展与应用,对领域内的主要研究方向和内容进行了分析和探讨,进而提出基于生物智能算法对群体机器人的控制方法进行优化。首先在菌群算法(Bacterial Chemotaxis)的基础上提出了改进菌群算法进行群体机器人自主目标搜索和围捕,利用自组织群体机器人解决该问题,提出了基于细菌趋化性的群体机器人智能控制算法实现目标搜索和围捕。根据目标区域内群体机器人的随机位置生成一个初始坐标系,再将目标区域划分为Voronoi单元,之后群体机器人在目标区域梯度信息的引导下实现目标搜索和围捕。仿真结果表明了算法的有效性及机器人发生故障时的鲁棒性。与群体机器人分布式控制等常用方法相比,模拟实验的结果表明,这种细菌趋局部最优,而且计算效率高。然后在上述算法的基础上,提出基于动态Voronoi图的群体机器人控制算法,在群体机器人进行任务操作时,根据机器人实时位置动态地划分目标区域,以确保机器人下一步的目标位置为距离自身最近的位置。仿真实验结果表明,算法的提出与实施有效地提高了群体机器人区域覆盖和目标围捕的效率,降低了重复探测率。接着,对于群体机器人控制中出现的带时滞随机连续噪声,给出基于随机基因调控网络(SGRN)和多目标优化的滤波器,仿真实验结果表明,滤波器的使用可以明显地提高在有噪声情况下基于SGRN群体机器人控制系统的收敛速度和运行精度。最后,对论文的全部工作进行了总结,并作出了群体机器人控制方法研究展望。
[Abstract]:With the development of robot application mode from component unit application to system application, robot application has gradually entered the public view. At the same time, more and more attention and research have been paid to swarm robot system because it can better meet the needs of people, and related research and applications have been springing up in endlessly. This paper introduces the origin, development and application of swarm robot technology, analyzes and discusses the main research directions and contents in the field, and then puts forward the optimization of the control method of swarm robot based on biological intelligent algorithm. Based on the bacterial colony algorithm (Bacterial Chemotaxis), an improved bacterial colony algorithm is proposed to search and capture autonomous targets of swarm robots, and self-organizing swarm robots are used to solve the problem. A swarm robot intelligent control algorithm based on bacterial chemotaxis is proposed to realize target search and capture. According to the random position of the swarm robot in the target area, an initial coordinate system is generated, then the target area is divided into Voronoi units, and then the swarm robot realizes the target search and capture under the guidance of the gradient information of the target area. The simulation results show the effectiveness of the algorithm and the robustness of the robot in the event of failure. Compared with the conventional methods such as distributed control of swarm robots, the simulation results show that the bacteria tend to be locally optimal and the computational efficiency is high. Then, on the basis of the above algorithms, a group robot control algorithm based on dynamic Voronoi diagram is proposed, which dynamically divides the target area according to the real-time position of the robot. To ensure that the robot's next target position is the nearest position to itself. The simulation results show that the proposed and implemented algorithm can effectively improve the efficiency of area coverage and target capture, and reduce the rate of repetition detection. Then, for the stochastic continuous noise with time delay, a filter based on stochastic gene control network (SGRN) and multi-objective optimization is given. The simulation results show that, The application of the filter can obviously improve the convergence speed and operation accuracy of the control system based on SGRN swarm robot in the presence of noise. Finally, the whole work of the thesis is summarized, and the research prospect of the control method of swarm robot is made.
【学位授予单位】：东华大学
【学位级别】：博士
【学位授予年份】：2016
【分类号】：TP242

【相似文献】