Abstract: Aiming at the defects of complex models, cumbersome operation and high cost of traditional energy consumption prediction methods for industrial robots, a meta-action-based energy consumption prediction method for industrial robots was proposed from the perspective of motion trajectory and action mode of end-effector. First, the meta-actions and energy consumption characteristics of static operations and dynamic operations of industrial robots were analyzed in the first place, and the energy consumption function of static meta-action and dynamic meta-action like axial motion and rotation et al, were built. Then, the motions of the target processes were disassembled based on the proposed meta-action library and the energy sonsumption calculation model of each meta-motion was built. Finally, energy consumption of industrial robots at target processes was realized by combining the energy consumption functions and calculation models. By applying the proposed method, the energy consumption prediction may be realized by disassembling operation processes as long as the energy consumption function of each meta-action was established by measuring meta-action power in advance, so the method has broad application prospects.

Key words:  , industrial robot, energy consumption prediction, motion path, meta-motion

摘要： 针对工业机器人传统能耗预测方法存在的模型复杂、操作繁琐、测量成本高等问题，从机器人末端动作模式和运动轨迹角度出发，提出了一种基于元动作的工业机器人能耗预测方法。首先分析工业机器人静态和动态运行时的元动作和能耗特点，建立轴移动、轴转动等动态元动作的能耗描述函数及静态元动作的能耗描述函数；然后基于所建元动作库对目标过程进行元动作分解，并构建各元动作的能耗计算模型；最后结合各元动作能耗描述函数和能耗计算模型，实现工业机器人目标过程的能耗预测。应用所提方法，只需事先测试元动作功率并构建能耗描述函数，就能通过分解目标过程实现能耗预测，具有较广阔的应用前景。

关键词: 工业机器人, 能耗预测, 运动轨迹, 元动作