农业轮式机器人PI鲁棒-滑模控制——基于RBF神经网络
|
摘要
农业轮式机器人机械多体系统朝柔性机器人方向发展,自由度越来越多,对应的结构也变得更加复杂,自动化和智能化水平越来越高,其动力学建模和实时控制难度增大。为提高机器人动力学建模效率,以通用性较强的具有6自由度机械臂的AMR果蔬收获机器人数学模型为研究对象,利用空间算子代数理论建立了轮式机器人O(n)阶效率的运动学和广义动力学模型。同时,利用Elman神经网络求解了机器人逆运动学问题,结合广义动力学模型和逆运动学模型,根据农业轮式机器人的特点,利用神经网络控制理论、PID鲁棒理论和Lyapunov稳定性理论,设计了一种6自由度机械臂的RBF-PI鲁棒-滑模控制算法,对机械臂末端进行心形轨迹实时追踪。最后,通过试验仿真,验证了本文提出的逆运动学理论、广义动力学模型和控制方法的合理性,为农业轮式机器人的研究提供了参考数据。
The rapid development of modern agriculture and task needs make the growth mechanical multibody system of agricultural wheeled robot towards Flexible robot.Their degree of freedom is increasing and corresponding system composition structure becomes more complex.In order to improve the efficiency of robot dynamics modeling,taking AMR xegetable and fruit automatic harvesting robot with 6 DOF manipulator for research objects,the o(n) hybrid dynamics of kinematics and generalized dynamics was gained by the spatial operator algebra theory.The problem of robot inverse kinematics is solved by elman neural network.With neural network theory,PID-Robust control theory and lyapunov stability theory,the RBF-PI-Robust-SMC control algorithm is designed to control the end manipulator to track the desired heart-shaped curve.The simulation results show that the given Inverse kinematics processing method,the generalized kinetic model and control method is logical.Data of the simulation provides more parameter date for the application of agricultural robot.
The rapid development of modern agriculture and task needs make the growth mechanical multibody system of agricultural wheeled robot towards Flexible robot.Their degree of freedom is increasing and corresponding system composition structure becomes more complex.In order to improve the efficiency of robot dynamics modeling,taking AMR xegetable and fruit automatic harvesting robot with 6 DOF manipulator for research objects,the o(n) hybrid dynamics of kinematics and generalized dynamics was gained by the spatial operator algebra theory.The problem of robot inverse kinematics is solved by elman neural network.With neural network theory,PID-Robust control theory and lyapunov stability theory,the RBF-PI-Robust-SMC control algorithm is designed to control the end manipulator to track the desired heart-shaped curve.The simulation results show that the given Inverse kinematics processing method,the generalized kinetic model and control method is logical.Data of the simulation provides more parameter date for the application of agricultural robot.
引文
[1]田富洋,吴洪涛,王冉冉,等.欠驱动空间柔性机器人动力学建模与实时仿真[J].力学与实践,2012,34(4):24-30.
[2]R K9ker,A genetic algorithm approach to a neural-network-based inverse kinematics solution of robotic manipulators based on error minimization[J].Elsevier Science Inc,2013,222(3):528-543.
[3]熊有伦.机器人技术基础[M].武汉:华中科技大学出版社,2014.
[4]曹东,田富洋.AMR果蔬自动收获机器人动力学建模与实时控制[J].农机化研究,2018,40(2):7-13.
[5]孙宏丽,吴洪涛,田富洋,等.空间算子代数刚柔耦合系统动力学软件流程[J].华中科技大学学报,2009,37(12):107-111.
[6]田富洋,吴洪涛,赵大旭,等.一类柔性宏刚性微空间机器人广义动力学建模研究[J].宇航学报,2010,31(3):687-694.
[7]Slotine Jean-Jacques E,Li Weiping.应用非线性控制[M].北京:机械工业出版社,2006:276-288.
[8]Ma Z,Sun G.Adaptive sliding mode control of tethered satellite deployment with input limitation[J].Acta Astronautica,2016,127:67-75.
[2]R K9ker,A genetic algorithm approach to a neural-network-based inverse kinematics solution of robotic manipulators based on error minimization[J].Elsevier Science Inc,2013,222(3):528-543.
[3]熊有伦.机器人技术基础[M].武汉:华中科技大学出版社,2014.
[4]曹东,田富洋.AMR果蔬自动收获机器人动力学建模与实时控制[J].农机化研究,2018,40(2):7-13.
[5]孙宏丽,吴洪涛,田富洋,等.空间算子代数刚柔耦合系统动力学软件流程[J].华中科技大学学报,2009,37(12):107-111.
[6]田富洋,吴洪涛,赵大旭,等.一类柔性宏刚性微空间机器人广义动力学建模研究[J].宇航学报,2010,31(3):687-694.
[7]Slotine Jean-Jacques E,Li Weiping.应用非线性控制[M].北京:机械工业出版社,2006:276-288.
[8]Ma Z,Sun G.Adaptive sliding mode control of tethered satellite deployment with input limitation[J].Acta Astronautica,2016,127:67-75.