Research On Fuzzy Logic-based Fault Synthetic Analysis Technique For NC Machine Tools

1.INTRODUCTION The fault Analysis of NC Machine Tools is not only important part of its reliability research, but also main chain of its reliability growth. Only Analyzing the fault of NC machine tools deeply, accurately and systematically can we find out the reason which caused fault, so that we can put forward improving measure to enhance its reliability. The conventional methods of fault analysis precisely describe randomicity of NC machine tools'fault, normal, reliability, fallibility and so on. These methods are based on a most basic hypothesis, which is 2-state hypothesis,namely the state of NC machine tools is normal or fault, no other state. As a matter of fact, on account of objective existence of illegibility, NC machine tools and its components can't be divided into normal or fault state simply. It is hard to evaluate the reliability of systems which are more and more tending to complicated. Furthermore, more intricate the system is, stronger its'fuzziness is, lower its precision is. The way to solve above problems is employing theory of fuzzy logic to analyze, evaluate and design. The fault synthetic technique based on fuzzy logic is given in this paper, which aims at the fault that the event is precise but probability is ambibuous. This method describes occurrence probability of fault events with fuzzy value, using fuzzy algebra operations to analyze all kinds of logic operators, thereby analyzing fault of NC machine tools. 2.FUZZY MATHEMATICS BASE OF FAULT SYNTHETIC ANALYSIS In this paper, fuzzy mathematics theory is applied to fault analysis methods of NC machine tools. In the course of fault synthetic analyzing, fuzziness of many factors are concerned, such as fuzzy occurrence probability of fault events (fault modes, basic events of fault tree etc.), fuzzy criticality of fault mode and so on. Besides various factors have different dimensions, membership functions of various factors and evaluation of membership value are different as well. Even as far as fuzzy occurrence probability is concerned, it is also likely to have many kinds of fault probability evaluations. The occurrence probability of fault events gained by various ways has many forms, including accurate occurrence probability value, language value and various fuzzy numbers. Therefore, for the sake of fault analysis, they should be unified. Because trapezoidal fuzzy numbers are membership function of linear distribution, algebra operations between them are simple, and translating other forms of fuzzy values into trapezoidal fuzzy numbers is easy and intuitionistic, so various fuzzy values are translated into trapezoidal fuzzy numbers. In the course of fault analyzing, algebra operations such as fuzzy operators of fuzzy numbers and relationship operations were use in this paper. 3.FAULT MODE, EFFECT AND CRITICALITY ANALYSIS BASED ON FUZZY LOGIC Put forward conception of fuzzy cognitive maps, conformed conception nodes. Conception nodes chosen in this paper included all the fault modes, sub-systems, fault reasons, fault reason classifications etc. Through analyzing field fault data came from a certain type of machining centers, fault effect propagation maps were set forth, which were established by fuzzy cognitive maps to show the relationship between fault modes and effect. The structure of machining centers are complex, and have too many conception nodes, so in terms of fault parts key conception nodes were defined, they were main spindle fault node, Z-axis feed system fault node, hydraulic system fault node, X-axis feed system fault node, Y-axis feed system fault node, tool magazine fault node, lubricate fault node, shield fault node, cooling system fault node, CNC fault node and other fault node. Fault effect propagation maps of these key nodes were established respectively. Analyzed the propagation tracks, indirect effect relation were analyzed by using fuzzy cause-effect algebra (min-max theorem). The way of fuzziness fault probability was bring forward, and calculating formula of fuzzy criticality was set forth likewise. This period of analysis had big utility for helping manufacturers to take improving measures, enhance unity reliability, maintain and preventively maintain. 4.FAULT TREE ANALYSIS BASED ON FUZZY LOGIC On the basis of FMECA, continuously processed fault analysis to the same machining tools. An available fuzzy fault tree analysis method was put forward. The structuring process of fuzzy fault tree was analyzed and explanation.