【作者】 朱齐丹；

【导师】 赵国良；

【作者基本信息】 哈尔滨工程大学 ， 控制理论与控制工程， 2002， 博士

【摘要】 具有非驱动关节机器人顾名思义，它不是所有关节都有驱动装置，也就是说，它有些关节是被动的，也称自由的。其特点是输入空间(即控制空间)维度小于构造空间维度，同时，自由关节的约束方程又是不可积的，具有非完全约束特性，这就增大了其在控制上的难度。这种机器人的优点是显而易见的，由于驱动的减少，使机器人的重量减轻了，成本也降低了，同时也就减少了能源的消耗，诸多优点使这种机器人很有研究价值，是近些年机器人研究领域的热点。 本文以该类机器人为对象，进行了位置控制的研究，主要完成以下几方面工作： 建立了具有非驱动关节机器人的动力学数学模型，对模型中的约束条件，即非驱动臂的动态方程的可积性进行了分析，给出可积性的判别方法，同时证明了该模型的约束条件为非完全约束。 通过相平面分析的方法分析了当驱动臂施加周期扰动对非驱动臂运动的影响，其结论为通过调整驱动臂周期扰动的幅度及极性，可以使非驱动臂沿着围绕平衡点±π／2的极限环轨迹运动。在此基础上，提出衰减因子的概念，通过衰减因子调整驱动臂扰动的幅度和周期，使非驱动臂能够稳定在平衡点±π／2。 以非驱动臂能够稳定在平衡点±π／2为基础，提出了一种对于具有非驱动关节机器人的基于神经网络的位置开环控制方法。利用前向BP网络能够对任意函数以任意精度逼近的特点，对开环控制策略中难以确定的函数进行学习，实现两关节的任意位置控制。 将驱动臂的运动作为非驱动臂的扰动，建立了非驱动臂的动态模型，利用动态方程有效的近似解法—平均法，对非驱动臂的动态模型进行简化，得到近似的平均系统，通过仿真计算验证了二者具有相同的动力学特性。 针对平均系统，利用李亚普诺夫稳定性方法设计了非驱动关节的位置闭环控制，使非驱动臂的运动从初始位置所在的极限环到达目标位置所在的极限环，为使非驱动臂能够稳定在目标位置，根据稳定目标位置速度为零的特点，推导出驱动臂扰动幅度在目标位置小邻域内的反馈规律，实现了非驱动臂的任意位置的臂环控制。 仿真结果表明本文设计的基于神经网络的位置开环控制方法及针对简化 哈尔滨工程大学博士学位论文后的平均系统设计的基于李亚普诺夫函数方法的非驱动臂任意位置臂环控制是非常有效的。更多还原

【Abstract】 Underactuated manimulator is a kind of manipulator with one or more joint is underactuated which can be called passive joint or free joint. The major character is that the dimension of input space is less than dimension of output space.We can see that the constraint of the system is un-integrable.It is called nonholonomic constraint.All of these feature make the control prolem has a number of additional difficulties.But there are many merit of this kind of manipulator,for example,underactuated manipulator maybe useful in particular instances to reduce weight,cost or energy consumption,while still maintaining an adequate degree of dexterity.So It is benefit to research the method of control problem,and it becomes one of the major topics of the field of robotics recently.The problem of position control is studied in this paper,and the results of my work are as follow:In first the dynamic model of underactuated manipulator is established,the non-holonomic characters of the constraint is analysed,and then the judgement method that if constraint is integrable is gived. So it is proved that the constraint of the manipulator in the paper is un-integrable,and the dynamic system is nonholonomic.Then the behavior of the free joint when we bring the periodic perturbation through controlling the actuated joint is analysed,We can find in the result that the behavior of the free joint in the phase plane follows an ordered closed trajectory with the center equilibrium point +2 in the case where the amplitude of the input is small,We can define a weakening factor to make the position of the free joint reach the center equilibrium point +2 .Based on the result above,the open loop control strategy is bringed forward in the paper. We use the function approximation property of neural network to obtain the function through studying using neural network method.And then realize the position control.We have simplified the model through obtaining the average system of theunderactuated robot using averaging method which is efficient to deal with nonlinear system. The simulation results show that the real system and the average system have the same trajectory in the phase plane under the disturbance of periodic input of the actuated joint of the robot.Finally,the closed loop control strategy through nonlinear states feedback is designed by a Lyapunov function for the averaged system and realize the position control of the non-holonomic robot.The simulation results show that the open-loop control method based on neural network and the feedback control method of the position are effective.更多还原