Research And Design Of Remote Monitoring And Diagnosis For Industrial Robot

Industrial robot is the key supporting equipment to promote manufacturing power,remote monitoring and fault diagnosis are the important guarantee to maintain its high reliability and high precision operation.The purpose of this paper is to solve the problem of remote monitoring and fault diagnosis of industrial robots under the conditions of significant individual differences and complex and changeable working conditions,and design an industrial robot remote monitoring and diagnosis system for mainstream equipment of major robot manufacturers.The design of the system includes three aspects: the design of abnormal monitoring method,the design of fault diagnosis method and the design of remote monitoring platform.Abnormal monitoring and fault diagnosis of industrial robots are very broad research fields.In this paper,due to the limited space,it is impossible to involve all aspects,so the collision detection of robots and the fault diagnosis of robot rolling bearings are selected for research,and the solutions are proposed(1)Aiming at the problem of collision detection of industrial robot,firstly,the dynamic model of robot is studied,and then the mechanical collision of robot is simulated by MATLAB.The research and simulation results show that the collision of robot will cause the sudden change of joint torque,this will cause the corresponding changes of joint angular acceleration and angular velocity.At present,a common collision detection method is to set the threshold of joint torque,which is prone to false positives and missing positives.In this paper,a robot collision detection method based on multi-dimensional parameter anomaly detection of robot is proposed.The method uses isolated forest algorithm to get the anomaly score of multiple parameters of each joint of robot,and then designs a decision rule to determine which link of robot collision occurs.(2)Aiming at the problem of fault diagnosis of rolling bearing of industrial robot,the structure and fault mechanism of rolling bearing of industrial robot are studied.It is found that the most obvious performance of rolling bearing fault is that the vibration intensifies,and the vibration frequency of different parts of rolling bearing fault is different.To solve the problem of bearing fault diagnosis of industrial robot,a rolling bearing fault diagnosis method based on vibration signal analysis is proposed.Firstly,the complete empirical mode decomposition method and wavelet packet threshold method are combined to denoise the bearing vibration signal.Then,the time-frequency characteristic image of the bearing vibration signal is obtained by wavelet transform.Finally,a deep convolution neural network is designed to classify the time-frequency characteristic image.Through the above three steps,the fault diagnosis of rolling bearing can be realized.In order to realize the remote monitoring of robot,a remote monitoring platform of industrial robot is designed based on B/S architecture.Through this platform,users can view the real-time running parameters and running status of robot anywhere with network.The platform development adopts the mode of front-end and back-end separation: the back-end is based on J2 EE platform and SSH framework,and deployed on tomcat7 server;The front end is developed based on Vue framework and element UI library.In the design of the platform,the following work has been done: in order to realize the data acquisition of different types of robots,the mainstream robot equipment interface has been investigated,and the data acquisition program has been written in the factory mode;In order to realize the secure transmission of data,a private data encryption transmission method is designed based on MQTT protocol;In order to ensure the fluency and ease of use of the system,AJAX technology or Web Socket protocol is used to realize the interaction between the end and the client.