The Research Of The Intelligent Virtual Maintenance Technology About The Wear Of Sliding Bearing Based On Internet

In the knowledge-based economic time, we will employ information technology to replace the natural resource in different extent, it has been fully visualized in the virtual development by the character of knowledge-based economic. In virtual production development process, virtual test is the essential point during the whole procedure, it is very convenient for us to use the research result of the actual device fault diagnosis to judge the virtual device fault diagnosis, predictive maintenance and the trace maintenance during the use. Generally the origin of the fault of the device is the sliding abrasion during the use of the device, so it is very important to do the research about the actual device sliding abrasion, which do not only have much value in practical use, but also provides a new way for virtual maintenance.In this paper, intelligence is firstly discussed, the computer network and virtual reality technology and its applications are also discussed. A framework of B/S based distant sliding bearing maintenance frame is given.In this paper, by using the system analysis method and analyzing the factor which plays an essential role in the metal sliding abrasion. Systemically giving the evaluating method and its relation with the maintenance ways. At the same time, introducing different methods for judging the abrasion status of Sliding Bearing. Here a Sliding Bearing had been as example to describe the life cycle of the sliding abrasion.Signal processing is important, traditional signal extraction can only extract and isolate the original signal. However the most unsatisfied with me is that they can not expand the extended information. Based on it, I provide a new idea about it called "the self-organized oriented computation network". It can make use of the relation among the different parameters to get the value of other parameters after the signal has been processed, naturally we canprovide enough information for the device fault diagnosis. The self-organized oriented computation network has many advantages. During the realization of the scientific computation, wee can also provide a way fitted to the habit of human being thought. This paper also gives an example.In this paper I make use of multiple recursive analysis and the optimal BP network to predict the friction and abrasion coefficient, by comparing and analysis, I get the reason "why the predictive result of Bayesian regularization is not better than the recursive predictive method, through deleting the " odd points ", the prediction result of Bayesian regularization method is improved largely , what is more the ANN can easily adapt the input values, so it is more robust. At last, a topology module is given with multiple input to predict the friction and abrasion coefficient of sliding bearing.At last, I also realize it using the VR technology aiding with Dynamic HTML technology and VC++. I use MATLAB to simulated the optimal ANN and embed it with Database Toolbox into the ASP to build a remote device maintenance systems. By doing this I realized a remote sliding bearing abrasion intelligent maintenance systems. In this system we calculate the abrasion and friction coefficient. We also predict the bearing life cycle at the same time giving the systems maintenance suggestion. From current result we got. The prediction result is better than traditional method.