An Ensemble Energy Consumption Prediction Model for Industrial Serial-Robot

Abstract

Industrial robots, as the representative of advanced technologies in multiple fields and disciplines, endow the manufacturing industry with automation and intelligence capabilities. Considering the sustainable manufacturing, it is essential to evaluate and optimize the energy consumption of industrial robots. State of the art research mainly focus on the energy consumption characteristic analysis,including the robot kinematics and dynamics.Parameter identification, the most crucial step, is difficult to implement for industrial circumstance. The traditional modeling methods reveals the non-linear relationship between the energy consumption and robot movements. This study uses the radial basis function neural networks to quantify the non-linear relationship. Moreover, an ensemble model method is implement to guarantee the accuracy of energy prediction. Experiments which demonstrates the effectiveness and accurateness of the proposed model are conducted on SiaSun SR4B industrial robot.