Research And Implementation Of Predictive Maintenance For Industrial Robot Based On OPC UA

The purpose of modern intelligent factory is to successfully integrate the traditional automated production workshop with the advanced technology elements of Industry 4.0to make the traditional factory more intelligent and digital.As an indispensable equipment in the construction of smart factories,industrial robots play a very important role in aviation,automobile,shipbuilding heavy industry and other manufacturing industries.In the actual production process,there are some problems such as high cost and lack of safety in manual maintenance of industrial robots.In this thesis,based on the traditional industrial robot welding system,a virtual simulation body is established to carry out predictive maintenance on the historical data of equipment,so as to realize the combination of virtual and real,and improve the predictive maintenance scheme of industrial robot.Thesis will carry out research work from the following aspects.1.The problem of equipment communication in predictive maintenance system of industrial robots is discussed.Thesis will study the communication between many devices and the virtual simulation body.OPC UA server is established on the basis of the traditional industrial robot and its auxiliary equipment,and the real device data is collected into the OPC UA server in real time for the virtual simulation body to access.2.For industrial robot movement process axial movement axis end failure time prediction problem,an improved grid search method of bidirectional LSTM algorithm,on the basis of the traditional grid search method selected suitable parameters to introduce improved the structure of the three layers of circulation search method to traverse the parameters to enhance model training precision lower loss rate prediction effect.The bidirectional LSTM algorithm based on the improved grid search method was used to predict the time of the repeated positioning accuracy deviation of the industrial robot in the fault period,and the maintenance strategy was proposed to realize the predictive maintenance.3.In order to solve the visualization problem of predictive maintenance system of industrial robots,thesis introduces a visualization method of virtual simulation body based on the traditional visualization method of predictive maintenance system of industrial robots.Traditional industrial robot predictive maintenance visualization methods are too single,thesis existing industrial robots and its subsidiary equipment CAD model for processing and import the Unity of script technology in 3 d form OPC UA client access OPC UA server real-time drive virtual scene simulation model and industrial equipment condition monitoring for the field devices simultaneously.