Research On Digitized Modeling Of Energy Consumption For Industrial Robots Driven By Physics-based Model

The global manufacturing industry is accelerating the steps forward to the goal of intellectualization at present,and intelligent manufacturing has become the common development direction of manufacturing industry all over the world.The unique characteristics of high efficiency,intelligence and high automation capability have made industrial robots increasingly used in various manufacturing process of manufacturing,it is the key equipment to support the development of intelligent manufacturing.However,the large-scale applications of industrial robots is often accompanied with high density and a large amount of energy consumption,at the same time,the demands for energy consumption reduction of industrial robots are getting higher because of the increasing environment and energy pressure and the upgrading of the manufacturing industry.The quantitative analysis and calculation of energy consumption of industrial robots is the basis to achieve the energy consumption reduction in the manufacturing process.Therefore,in order to promote the energy saving in the manufacturing industry,the effective modeling,simulation and prediction for the energy consumption of industrial robots is extremely necessary.In this paper,the serial industrial robots that are commonly used in the manufacturing process are taken as research object,and the physics-based energy model of industrial robots based on inertial parameters identification is established firstly,combined with the theories and methods of digitized description and virtualization modeling,a digitized energy modeling method driven by the physics-based energy model of industrial robots is proposed.The main work of this research is as follows:(1)A physics-based energy model of industrial robots based on dynamic parameters identification using the Gbest-guided artificial bee colony algorithm is proposed.Considering that the energy consumption of industrial robots in the running process is mainly caused by joint motor and friction loss,the dynamic equation of industrial robots are obtained firstly according to the robotic kinematic and dynamic theory.The application of the Gbest-guided artificial bee colony algorithm is proposed to identify the unknown parameters in the equation to get higher accuracy.On this basis,the dynamic equation is then applied to the energy calculation of industrial robots and the effectiveness of the result is validated,thus the physics-based energy model is established.(2)Research on the digitized energy modeling of industrial robot driven by physics-based energy model.Aiming at the knowledge description and visualized interaction mechanism for energy consumption of industrial robots in the manufacturing process,the OWL ontology language is utilized to build the digitized description model in the domain of energy consumption of industrial robots,the mapping relationships between the data of physics-based energy model and the ontology attributes are excavated and established,inference rules are built to realize the knowledge reasoning of the ontology model based on the data driven from the physics-based energy model.Then,the visualized and digitized virtual modeling for the energy consumption of industrial robots is carried out.On this basis,a integration model which associates the physics-based model,digitized description model and digitized virtualization model is obtained to realize the interrelation of physics-based and digitized model and the fusion of the data and knowledge.(3)A prototype system for energy analysis and prediction of industrial robots is designed and developed.With the data-driven by the physics-based energy model,the knowledge-driven by the digitized description model and the visualization and interaction capability provided by the digitized virtualization model,the effective monitoring,simulation and prediction of energy consumption in the manufacturing process of industrial robots can be achieved to support further energy reduction and optimization.