（毕业论文）机械手轨迹跟踪智能控制研究.doc

摘 要 智能控制理论是具有某些仿人智能的工程控制和信息处理方法。机器人学集中了机械工程、电子工程、计算机工程、自动控制工程以及人工智能等多种科学的最新科研成果，是目前科技发展最活跃的领域之一。 机器人的轨迹控制在机器人的控制中具有重要的地位。论文首先回顾了智能控制技术及其在机器人控制中应用的发展和现状。详细介绍了机器人的数学模型。在分析了模糊神经网络的各种学习算法的优缺点后，研究了一种新的轨迹优化算法。给出了机械手动态模糊神经网络控制的基本机构和算法，设计了一个性能优良的机械手轨迹跟踪控制系统。以SEIKO TT-3000机器人为研究对象，建立了机器人跟踪数学模型。基于T-S 模型设计出MIMO的RBF 模糊神经网络结构,给出RBF模糊神经网络的基本算法，设计了RBF模糊神经网络控制器。在此基础上将RBF神经网络与TSK模型相结合，设计出动态模糊神经网络结构，并给出其基本算法，设计了机器人操作臂的控制方案。 利用虚拟样机技术，对机器人手臂控制用Matlab/Simulink进行仿真，在无干扰和有干扰两种情况下，给出了轨迹跟踪结果。研究机械手的运动学和动力学指标在运动过程中的变化情况，验证了该优化算法的可行性和先进性。仿真结果表明该算法实现了轨迹规划的快速收敛和近似逼近，具有一定的容错能力，避免了因约束点的输入顺序而产生的影响。 关键词 机械手；跟踪控制；动态模糊神经网络控制 Abstract Intelligent control theory is theory methods with some imitate people project control and information processing．Robotics is the newst scientific findingshas concentrated many kinds of disciplines,such as mechanical engineering,electronic engineering,computer project and artificial intelligence,etc.,it is one of the most active fields of the development in science and technology at present． Robot’s trajectory planning is very important in the control of robotics．At first, this thesis reviews the intelligent control technology and its application in robot control and development status．The mathematical model of the robot is introduced in detail．After analyzing the advantage and disadvantage of various kinds of learning algorithm of the FNN, a new trajectory optimization of simulation is．Given manipulator dynamic fuzzy neural network control of the basic structures and algorithms, and tries to design a good control system for the rigidity robot manipulatorIt introduces dynamic neural network theory and its basic algorithm with the SEIKO TT-3000 robot as research object, establishes the mathematical model of the robot tracking, designed MIMO RBF Fuzzy Neural Network Structure and it based on TS model, this paper gave RBF fuzzy neural network algorithm,and designed RBF Fuzzy Neural Network Controller．I combine RBF neural network with the TSK model combining and the