自平衡机器人的移动稳定性与路径优化研究

发布时间：2019-06-11 09:59

【摘要】：近年来,移动机器人技术水平有着飞跃式的发展,不仅在生产生活中被广泛应用,而且对许多国家工业、国防以及国民经济产生了巨大的影响。它在实际应用中可代替人工在危险(有毒、辐射等)、复杂环境中作业,可从事在人所不及(太空、深海等)的空间环境中。其中两轮自平衡机器人是移动机器人中应用较为广泛的一种,它是集环境感知、动态决策与规划等多功能于一体的智能化机器系统,它融合了多级倒立摆的不稳定性、多变量、非线性和强耦合等特点。由于两轮机器人可以实现多轮机器人无法实现的复杂动作,因此对它们的深入研究具有重要的理论及现实意义。本文针对自平衡机器人在实际应用中存在稳定性不佳、人工经验选择控制器权值结果较差等问题,展开了系统研究,在基于线性二次性最优控制器和自平衡机器人模型的基础上,提出了一个基于蚁群算法优化机器人稳定性的模型。之后,在保证了机器人安全性、平稳性的基础上,针对机器人路径规划中前期环境地图信息采用划分方式不理想和后期单一算法优化路径不佳等问题,提出一种改进蜂巢栅格法来优化前期环境信息并采用一种基于遗传-蚁群的混合算法来解决后期最优路径的问题,使系统平稳且无碰撞的完成路径寻优。本文研究工作主要包括:(1)本文提出一个基于蚁群算法优化机器人稳定性的模型,用以选择适合控制器的参数权值,克服了经验选择参数的不确定性和耗时性。此模型对最优控制器的性能指标函数对于状态量的权阵参数进行优化,通过最优的权阵结果得到控制器解值,最终决定移动机器人的稳定指标结果。该优化方式更优于传统人工选择参数的方法,不仅优化了机器人的稳定性,还降低了实践中相关系统应用的危险系数。(2)本文提出一种改进蜂巢栅格法来处理路径优化前期环境信息,它结合了自然生物现象和几何学理论,克服了传统栅格法在处理地图信息时占用过多可行空间、增加机器人碰撞率的缺点,并且改善了蜂巢栅格法编码时不连续、不能根据起始点和终点的位置控制编码规则的问题。该方法采用自适应编码方式,灵活操作地图信息的设定,使栅格具有连续性,更利于最优路径的搜索,有效的减少机器人行进的路径长度和路径搜索时间,提高机器人的安全性。(3)本文在处理路径规划后期搜索最优路径的问题中,选定蚁群和遗传算法作为混合方法的研究基础,通过实验对比分析,结合二者的优缺点,采用了一种基于遗传-蚁群的混合算法对路径进行寻优,克服了传统路径规划采用单一优化算法搜索路径时稳态迭代次数和稳态时路径长度不理想的问题,并通过设置30次、80次和150次的迭代次数,可以得到多混合算法均优于单一算法的处理结果。
[Abstract]:In recent years, the level of mobile robot technology has been developed by leaps and bounds, which is not only widely used in production and life, but also has a great impact on industry, national defense and national economy in many countries. It can be used instead of artificial in dangerous (toxic, radiation, etc.), complex environment, can be engaged in the space environment (space, deep sea, etc.). Two-wheeled self-balancing robot is widely used in mobile robots. It is an intelligent machine system which integrates environmental perception, dynamic decision-making and planning. It combines the instability and multivariable of multi-stage inverted pendulum. Nonlinear and strong coupling and other characteristics. Because the two-wheeled robot can realize the complex action that the multi-wheeled robot can not realize, it is of great theoretical and practical significance to study them deeply. In this paper, a systematic study is carried out to solve the problems of poor stability and poor weight value of artificial experience selection controller in the practical application of self-balancing robot. Based on the linear quadratic optimal controller and the self-balancing robot model, a model based on ant colony algorithm is proposed to optimize the stability of the robot. After that, on the basis of ensuring the safety and stability of the robot, the environmental map information in the early stage of robot path planning is not well divided and the path optimization by a single algorithm in the later stage is not good. An improved beehive grid method is proposed to optimize the environmental information in the early stage and a hybrid algorithm based on genetic ant colony is used to solve the problem of optimal path in the later stage, which makes the system stable and collision-free. The main research work of this paper is as follows: (1) in this paper, a model based on ant colony algorithm to optimize the stability of robot is proposed, which can select the parameter weights suitable for the controller and overcome the uncertainty and consumption of empirical selection parameters. In this model, the performance index function of the optimal controller is optimized for the weight matrix parameters of the state quantity, and the solution value of the controller is obtained by the optimal weight matrix result, and finally the stability index result of the mobile robot is determined. This optimization method is better than the traditional manual parameter selection method, and not only optimizes the stability of the robot. It also reduces the risk coefficient of the application of related systems in practice. (2) in this paper, an improved hive grid method is proposed to deal with the environmental information in the early stage of path optimization, which combines natural biological phenomena and geometry theory. The shortcomings of the traditional grid method in dealing with map information are overcome, such as taking up too much feasible space and increasing the collision rate of the robot, and the problem of discontinuity in the coding of the hive grid method is improved, and the coding rules can not be controlled according to the position of the starting point and the end point. In this method, the adaptive coding method is adopted to flexibly operate the setting of map information, so that the grid has continuity, which is more conducive to the search of the optimal path, and effectively reduces the path length and path search time of the robot. Improve the safety of the robot. (3) in dealing with the problem of searching the optimal path in the later stage of path planning, ant colony and genetic algorithm are selected as the research basis of the hybrid method, and the advantages and disadvantages of the two methods are compared and analyzed by experiments. A hybrid algorithm based on genetic ant colony is used to optimize the path, which overcomes the problem that the number of steady-state iterations and the path length in steady state are not ideal when the traditional path planning uses a single optimization algorithm to search the path, and by setting 30 times, With 80 iterations and 150 iterations, the processing results of multi-hybrid algorithm are better than those of single algorithm.
【学位授予单位】：昆明理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242

【参考文献】