面向电子装配机器人高速精密运动实现关键技术

Integrated circuit assembling robots are required to implement high-speed precise motion, and the precision is 1-2 orders of magnitude higher than traditional industrial robots (100μm). In high-speed operating process, it is difficult to damping the inertia energy in an extremely short period, which will bring challenge to precise positioning. When mechanism motion comes to ‘high-speed’ domain, the moving components mush be treated as flexible bodies. Therefore, the dynamic problem will be modeled as high-dimension differential equations containing terms with variable coefficients, non-smooth and nonlinear or even strong nonlinear, which will bring difficulty for modelling, analysis, optimization and control. Based on the features of rigid-flexible coupling dynamic model, and considering from the aspect of subjective and passive vibration control, this project condenses two main scientific problems for high-speed mechanism, which are reduction and compensation of rigid-flexible coupling, multi-excitation and multi-mode superposition vibration. For vibration reduction, novel methods are proposed for structure design and motion planning based on optimal inertia energy space-time distribution. For vibration compensation, a multi-mode vibration compensation method is developed based on macro-micro compound design. This project are making efforts to solve the space-time discretization, modal equivalence, modal separation, positioning accuracy determination, macro-micro design and macro-micro control problems of dynamic response equations with rigid-flexible coupling, nonlinear time-varying, multi-excitation and multi-mode superposition. The research of this project can provide theoretical foundation and method supporting for vibration reduction and compensation of high-speed precise robots.

电子装配等领域的机器人要求实现高速精密操作，其精度比传统工业机器人(100um)高1-2数量级。在高速操作时，惯性能量难于在极短时间内衰减，给精密定位带来挑战。当机构运动进入“高速”区域时，运动部件必须作柔性体假设，动力学模型将以变系数、非光滑、多非线性甚至强非线性项组合的高维微分方程组形式出现，给建模分析、优化和控制带来困难。本项目基于刚柔耦合非线性动力学模型特征，从主被动振动控制角度，凝练了高速机构刚柔耦合多激励多模态叠加振动的抑制和补偿两个主要科学问题。在振动抑制方面，建立基于惯性能时空最优分布的结构设计和运动规划新方法。在振动补偿方面，建立基于宏微复合设计与控制多模态振动补偿方法。主要解决刚柔耦合非线性时变多激励多模态叠加动力学响应方程的时空离散、模态等效、模态分离、定位精度判定、宏微复合设计和宏微复合控制问题。本项目的研究将为高速精密机器人的振动抑制和补偿提供理论依据和方法支撑。

电子装配等领域的机器人要求实现高速精密操作，其精度比传统工业机器人(100um)高1-2数量级。在高速操作时，惯性能量难于在极短时间内衰减，给精密定位带来挑战。当机构运动进入“高速”区域时，运动部件必须作柔性体假设，动力学模型将以变系数、非光滑、多非线性甚至强非线性项组合的高维微分方程组形式出现，给建模分析、优化和控制带来困难。本项目基于刚柔耦合非线性动力学模型特征，从主被动振动控制角度，凝练了高速机构刚柔耦合多激励多模态叠加振动的抑制和补偿两个主要科学问题。. 在振动抑制方面，建立基于惯性能时空最优分布的结构设计和运动规划新方法。 在惯性能空域优化即柔性多体动力学优化方面，提出了基于重分析的快速求解方法和基于梯度投影法和Kiriging代理型的复合全局优化方法，获得的更优的结构性能。在惯性能时域优化即运动规划方面，提出了驱动与抑振融合的于运动规划方法，实现了谐振的避免和运动过程振动分量的抵消，避免了输入整形对改变运动轨迹和执行时间滞后。.在振动补偿方面，发明了刚柔耦合平台结构与刚柔耦合旋转关节，采用柔性铰链补偿摩擦死区，提高精度。将摩擦力扰动转换为弹性力扰动，并通过自抗扰控制算法的扩张状态观测器，估计参数和载荷扰动，实现机器人与环境的共融。开发出的刚柔耦合平台，重复定位精度达到纳米级。