摘要
针对机械臂自适应神经网络控制效果不能直观显示的问题,在深入研究各种虚拟仿真,可视化仿真方法的基础上提出了一种基于ADAMS和MATLAB的虚拟样机联合仿真方法。采用该方法对三自由度机械臂进行动态仿真,首先在ADAMS中建立机械臂模型,然后在MATLAB中实现自适应神经网络控制器和机械臂终端轨迹的设计,最后进行联合仿真。仿真结果非常直观地显示了机械臂的动态控制过程和控制器各参数的实时变化情况,验证了该方法的有效性,为机械臂的复杂控制算法的仿真研究提供了新的有效方法。
Considering the problem that the adaptive neural control of robot manipulators can not show the control result obviously,a new co-simulation method based on ADAMS and MATLAB is proposed through analyzing the method of virtue simulation and visual simulation.In order to show the efficiency of the method,the 3-dof robot manipulator is chosen as the control object.Firstly,a virtual modeling of robot manipulators is created by using ADAMS software.Then,the adaptive neural control algorithm and the terminal trajectory of the robot manipulator are implemented in MATLAB.Lastly,a Co-simulation between ADAMS and MATLAB is realized.The simulation results indicate that the control process of the robot manipulator and the parameters of the controller can be watched intuitively witch show the effectiveness of the method.
出处
《组合机床与自动化加工技术》
北大核心
2012年第3期50-53,56,共5页
Modular Machine Tool & Automatic Manufacturing Technique
基金
国家自然基金(61075082)
关键词
机械臂
神经网络控制
虚拟实现
联合仿真
robot manipulators
neural network control
virtual realizing
co-simulation
