柔性机械臂的建模与控制研究.pdf.pdf

摘要本文以两连杆柔性机械臂物理模型为研究对象,预测神经网络控制系统为研究框架,建立了柔性臂的动力学方程及预测神经网络控制器,主要的内容如下: 第一,柔性臂动力学方程的建立及其振动响应仿真。对柔性机械臂的物理模型进行简化,将柔性臂视为Euler-Bernoulli梁,并采用假设模态法与Lagrange方程推导出柔性臂的动力学方程。。第二,建立BP神经网络辨识模型。构建BP神经辨识网络,并以振动分析得到的数据对网络进行训练和测试,最终得到控制系统中的BP神经网络预测器。第三,刚性机械臂的轨迹规划。以臂杆角速度为最优化控制优化目标,采用中心差分法,逐步迭代出刚性机械臂的理想轨迹,这将作为控制器的期望轨迹。最后,柔性机械臂控制器的综合设计。将BP神经网络预测器、期望轨迹和神经网络控制器结合组成神经网络预测控制系统。通过仿真,表明所设计神经网络预测控制系统可实现柔性臂按规划轨迹平稳运行的目标。关键词:柔性机械臂轨迹规划神经网络预测控制 Abstract Based on aphysical model of a two-link flexiblemanipulator,a study on the modeling andcon们l offlexiblemanipulator,using the work predictive corarol,is preserted inthepaper. Firstly,the dynamic equatiom ofthe flexiblemanipulator are physicalmodel oftheflexiblemanipulator isreduced bytaking theflexiblelink嬲 Euler-Bernoulli dynamic equations ofthe flexible manipulator are developed by theLagrange equationandthe aSsumed mode thevibration analysis obtained by theNewrmrk一13 method wellverifes tracy ofthemodel Secondly,the work identification model is aBP work identificationsystem,and using the datafromthe vibration alialysis to trahand testit,thus we can get workpredictor. Thirdly,the trajectory of rigid flexiblemanipulator the Optimization method withthevebcity ofthe arnl嬲its optimization goal and the centraldifferencemethod toplanthe idealtrajectory of atwo-link rigidrmnipuhtor, thisisthedesiredtrajectory ofthe eorm-oller. Finally,the controllerofthe flexiblemanipulator BP work predictor,ideal trajectory and workcontroller bined tobe workpredictive simulation shows that theflexiblemanipulator can move smoothly according tothedesndtrajectory under thecorm'olofthe work predictive corarol sys:tem. Keywords:Flexible Manipulator Trajectory Planning work Predictive Control 第一章绪论第一章绪论 、单调及乏味的工作ll】。然而, 现如今大多数的机械手臂都是传统的刚性机械手臂,这些传统的机械手臂大多是由重型的材料设计而来,这种笨重的设计不可避免的导致了刚性机械手臂运行中的负载自重比低、能耗大、响应速度慢、定位精度低等诸多问题。