Research On Remote Fault Diagnosis And Control System For CNC

With the rapid development of computer network technology, CNC technologyconstantly uses the computer, modern control theory and other areas of the latest scientificand technical achievements towards the running speed,intelligent control,precision machiningtechnology, parallel drive, interactive network, greening, openness and the direction of theflexibility. Advanced CNC technology has led to the rapid development of the manufacturingsector, the growing number of CNC machine tools, and the CNC equipment is more and morecomplicated, and the higher degree of integration and automation, making the increased fault.At the same time, because improper use or fatigue work also increase the increased failurerate of the equipment, If the manufacturing enterprise can not provide timely maintenancewith end-users of CNC machine tools, that will affect their production plan, and bring hugeeconomic losses. It is essential to provide good technical support for its sold products whichcan solve fault, improve the quality of after-sales service, and adapt to the market futuredevelopment needs. In the network times of highly developed information, the remotediagnosis technology can be a very good way to solve the above problems.This topic comes from the special projects of research and development departmentunder taken by Baolun CNC Technology Co., Ltd.–“the remote diagnosis and control forCNC”. Firstly, this paper introduces the BaoLun’s CNC system structure as well as remotediagnosis technology’s research situation of the domestic and international; Secondly,thepaper introduces the structure,principle,the advantages and disadvantages of the two networkmodel about the C/S and B/S, and ensure the network architecture of this paper, meanwile,the paper also introduce the principle about data transmission; then, the paper simpleillustrates several methods about data analysis, the selecting data analysis method--theHilbert-Huang transform is introduced detailed which includes EMD decompositionalgorithm, the IMF component, and the marginal spectrum by Hilbert transform for everyIMF components. Finally, pick up the characteristics of the fault data.By the characteristicswe can also identify some fault information from the marginal spectrum; Followed by,analysis the methods of fault diagnosis module detailed, which including expert system andneural network’s structure, principle, advantages and disadvantages, and ultimately determineusing the combination of the two as this paper fault diagnosis module diagnosis methods;Finally, in the Windows XP operating system platform, using Visual C++6.0, MATLAB,ACESS and other development tools,combined with ADO object, ActiveX controls andTCP/IP agreement, WinSocket socket technology, and object-oriented method, develop theremote fault diagnosis system software, which users can through the man-machine interface to achieve fault information query, network communication, fault diagnosis. In addition, this textsimulates the improved EMD and BP neural network which applied in this topic in chaptersthree and chapters four, through the EMD algorithm extractes the characteristics of faultsignal, and then uses the BP network training, and identify the fault successfuly. The end ofthe article, summarize the full text, point the shortcoming,throw out the improved advice andput forward recommendations on the issue needs further study.The results of this paper research greatly improve the management level of the company’safter-sales service, improve the competitiveness of their products, and make the CNCequipment adapt to the future development needs of the market. It is a very active role in thelong-term development of the enterprise.