The Intelligent Compensation Method And System Design Of Thermal Deformation Of The Numerical Control Machine Tools

As we all know, thermal error is a crucial factor in machining process, which influences the accuracy of the NC machine. In total machining error,The proportion of thermal error is up to70%-80%[1].So how to reduce the thermal error is so important. To improve the precision of the NC machine to reduce the thermal error, we should take out a NC machine tool error compensation technology, which should be an effective method and means to improve the accuracy of it. In recent years, Although the scholars made some progress, the problem of Recognition and Management of Original error and establishing the compensation model of thermal error have not been solved effectively. How to determine the temperature-sensitive point location and establishing the compensation model of thermal error and software application in thermal error real-time compensation are the two crucial problems in this paper.In this paper, We take the research combining with Artificial Intelligence Key Laboratory of Sichuan Province project "the key technology and application research of the Open mode intelligent NC system"(2010RZ002) and SiChuan ChangZheng Machine Tool Group Major national research projects "high-speed longmen five axis machining center"in2009(2009ZX04002-013).The main contents of the thesis are as follows:(1) In order to establish the compensation model of thermal error, it introduces the basic knowledge and concept of Rough Set Theory and Artificial Neural Networks, it analysis the source of thermal error and the thermal error cause, it also introduces the Identification and measurement methods of thermal error.Put forward a new method of thermal error measurement with Rough Set Theory.Put forward a new method to reduce the thermal error.(2) For the temperature-sensitive point location.Analysis the problems of Common method.Put forward a new method to determine the temperature-sensitive point location with Rough Set Theory.(3) For the establishing the compensation model of thermal error. Analysis the problems of Common method establish a compensation model of thermal error with BP neural net work. Put forward a Gradient descent training online learning algorithm.This simple and effective algorithm has less Computation and short Network training time which improved the performance of BP network.(4) Combining with GM2000A NC machine tools to Analysis Temperature Measurement Point and establish the compensation model of thermal error.the simulation results prove it is feasible.(5) For the maching Change some PLC Programs to shift the prokinematic pair to produce a relative motion in the inverse direction of the spatial error of the machine. Practice has proved that this intelligent compensation is an Effective and Practical method.Innovative contribution in this article is as follows:Put forward a new method of thermal error measurement with Rough Set Theory.It optimizations the Temperature Measurement Point with Rough Set Theory. It simplifies the inputs of the BP neural net work, a gradient descent training online learning algorithm Optimizations the net. Improve the Efficiency and Prediction accuracy of the net. At the same time, Editor compensation programs with S7-300PLC software. It compensates the thermal error Effectively, Improved the machining accuracy and reduced production costs...