多关节致动器非线性摩擦最小能耗控制模型研究

Reducing the energy consumption of industrial robots is one of focus in the development of green intelligent manufacturing system. The energy efficiency of industrial robots is inhibited obviously due to the friction loss of joint actuators. In this project, we focus on optimization strategy of the controlled electromechanical loss for the multi-joint manipulator on heavy load and high speed operation, which based on the mechanism of the nonlinear friction on macro and micro scale. This research needs a foundational method to design and realize the tests for controlling dissipated losses of the joints. It is combined with our previous results. That is the off-line programming and precision interpolation for motion trajectory of the end actuator. Through of these works, we will set up a new electromechanical coupling dynamic model of multi-joint actuators with nonlinear friction. To explore the general method to obtain the optimal control current and the optimal velocity function in PTP and CP mode. Meanwhile, we will propose and verify a self-adaptive control algorithm for the optimal energy trajectory of robotic manipulator on heavy load and high speed operation, when the prediction model of the minimum energy is established. All of efforts will provide a new theory and method to design and control the energy saving joint actuator. The results has important influence on the energy efficiency improvement of industrial robots.

工业机器人的节能降耗是绿色智能制造系统发展中的焦点问题之一。关节致动器的摩擦损耗对工业机器人的能源效率有明显抑制作用。结合我们前期对末端执行器运动轨迹离线编程精准插值问题的阶段性探索成果，基于关节能耗控制试验的设计与实现，探析宏微观尺度下非线性摩擦的作用机理，重点研究重载、高速运行时多关节机械臂可控机电损耗的优化策略。力图建构一种新的多关节致动器非线性摩擦机电耦合动力学模型；探寻PTP、CP模式下最优控制电流和最佳速度函数的一般性获取方法；通过多关节致动器最小能耗预测模型的建立，系统提出重载、高速运行时机械臂最优能量轨迹的自适应控制算法，并验证其有效性。可为节能型关节致动器的设计与控制提供新的理论、方法支持，对工业机器人能源效率的提升具有重要意义。

基于关节能耗控制试验的设计与实现，综合考虑电机热损耗、关节摩擦损耗和机械运动能耗，建立起高速运行机械臂的运动学模型、刚体动力学模型和考虑关节摩擦的动力学模型，进而提出一种基于能量模型的关节摩擦辨识新方法；系统建立负载转矩约束下的重载搬运机械臂驱动系统机电耦合动力学模型，推导出非线性摩擦作用时的最优角速度和最佳控制电流表达式，提出一种基于改进BP神经网络算法的关节最小能耗软测量模型，可实现未知角位移和未知负载情况下关节能耗的准确预测。运用含约束条件的遗传算法达成高速运行机械臂的五次非均匀B样条最优能量轨迹规划；改进的五段S型曲线插值算法解决了重载机械臂关节角加速度突变及波动范围大的问题。所提出的力矩前馈控制策略可较为精准地驱动机械臂各运动关节沿预定轨迹运动，且动态性能指标稳定、能量耗散较小，为节能型关节驱动控制系统的设计提供了新的理论和方法支持。项目执行期间，在国内核心期刊发表论文6篇，在国际学术会议发表论文2篇、已录用待发表1篇，其中1篇被SCI收录、2篇被EI收录；培养硕士研究生6名。本项目投入专项经费43.2万元，支出24.2419万元，各项支出基本与预算相符。剩余经费18.9581万元，计划用于本项目研究后续支出。