RV Reducer Health Status Evaluation Method Based On Optimized Support Vector Machine And Improved Evidence Theory

With the transition from manufacturing in China to intelligent manufacturing in Ch ina,industrial robots play an important role as important equipment for intelligent manu facturing,and joint reducers,as the core components of industrial robots,play an import ant role in the precise motion control of robots.Among various joint reducers,the RV re ducer has become the mainstream joint reducer for industrial robots due to its advantage s such as high transmission ratio,high transmission stiffness,high transmission accurac y,smooth motion,and small backlash.At the same time,the health status of RV reducer s has an important impact on the health status of industrial robots.Accurate evaluation o f the health status of RV reducers and formulation of appropriate maintenance strategies are crucial to improving the reliability level of industrial robots.This paper takes the RV-80 E reducer produced by a company as the research objec t,builds a reliability test system for RV reducers,collects test data,and proposes a healt h status assessment method based on single signal and multi signal information fusion.T he main research work is as follows:(1)The typical failure modes of the RV reducer are analyzed.The operating princip le and load conditions during operation of the RV reducer are introduced.The T-S fuzzy fault tree of the RV reducer is established.The failure probability of the entire machine is calculated based on the failure probability of each subcomponent,and the degradation of the entire machine under different degradation conditions is calculated for each subc omponent of the RV reducer,This provides a basis for the subsequent selection of key s ubcomponents of the RV reducer and their corresponding health status monitoring indic ators.(2)A reliability test system and a condition monitoring system for electrically close d RV reducers based on motor coupled simulation loading were established.Referring t o the rated working conditions of the RV reducer,a reliability test system for electrically closed RV reducers was established using servo motor coupling to simulate loading,an d a PLC loading control program based on TIA Portal was developed.According to the fault analysis results of the RV reducer,the vibration,temperature,output shaft circular runout,servo motor current and other signals of the RV reducer are selected as monitori ng indicators,and a RV reducer status monitoring system based on Labview is establish ed.(3)A single signal based health assessment method for RV reducers is proposed.A single signal based health assessment method for RV reducers is proposed and validated using vibration signals as an example.By applying different eccentric loads to the outp ut end of the RV reducer to simulate the different health states of the RV reducer,the ax ial vibration signals at the needle gear housing of the RV reducer are collected,grouped and processed,and subjected to 3-layer wavelet packet decomposition.Then,the range e ntropy of each decomposition vector is calculated to construct an 8-dimensional feature vector,which is input to SSA-LIBSVM for training and classification.(4)A method for evaluating the health status of RV reducers based on information f usion is proposed.A method for evaluating the health status of RV reducers based on m ultisensor information fusion is proposed,and the evaluation is validated using multi-dir ectional vibration signals and output shaft circular runout signals as examples.The axial vibration signal,radial vibration signal,circumferential vibration signal,and output sha ft circular runout signal of the RV reducer that have been verified to have good classific ation effects are selected as the input signals for health assessment of multisensor fusion,and their respective evaluation results,namely a posterior probability distribution,are o btained through a single signal evaluation method.On this basis,each sensor is subjecte d to subjective weighting based on BWM and objective weighting based on Jousselme e vidence distance.The optimal allocation proportion of subjective and objective weights is obtained through SSA optimization algorithm for combined weighting,and then D-S evidence theory fusion is performed on them to obtain the final evaluation result.