Research On Reliability Evaluation Of Rice Transplanter

Rice is the food crop with the widest planting area and the largest total output in China,and the rice transplanter is the main machine for transplanting rice seedlings.In recent years,with the development of agricultural mechanization in China,rice transplanters have been put into use in many plain areas.During the period of transplanting rice seedlings,the operation time is limited and the labor intensity is large.The reliability of the rice transplanter is very important.Its reliability is not only related to itself,but also related to the maintenance cost of the machine and the loss caused by the failure to transplant seedlings on time due to the shutdown.Therefore,it is very important to study the reliability of rice transplanter.Reliability evaluation is an important research content of rice transplanter.The true level of reliability can be obtained by reliability evaluation.The multiple fault parts and fault reasons can be found out.The results of reliability evaluation can provide a detailed basis for improving and enhancing the reliability of rice transplanter.In view of this,this paper selects a certain type of rice transplanter as the research object,and conducts an in-depth reliability evaluation study.The main research contents are as follows:(1)Reliability test of rice transplanter and FMEA.According to the reliability data and failure characteristics,the sub-parts,failure mode and fault reason of the rice transplanter system are divided.Then,the FMEA was used to analyze the whole rice transplanter and its multiple fault sub-parts.(2)Reliability evaluation of rice transplanter based on mixed weibull distribution.Firstly,a mixed weibull distribution is selected as the reliability model for the rice transplanter which has multi-failure modes and complex fault mechanism.Secondly,the fault stress similarity is proposed to characterize the fault mechanism similarity.The expert evaluation method is applied to establish the relationship between the failure phenomenon and the failure stress.The incidence matrix between the failure mode and the failure stress is builed through the specific failure phenomenon.Then,the fuzzy cluster analysis is used to divide the failure modes into two categories,and a mixed double weibull distribution model is builded.The model parameters are estimated by nonlinear least square method,and the goodness of fit test is carried out.Finally,the reliability evaluation result of rice transplanter is obtained by the model.(3)Reliability evaluation of rice transplanter based on neural network.First,the reliability models of BPNN and RBFNN are builed.A better performance network is selected by comparison,and the reliability index of the rice transplanter is obtained by using this network model.Then,In order to solve the problem of less original sample data,RBFNN is used to build an extended model of reliability data.The large sample data with the same failure statistical rule as the original sample is obtained by extension model.A mixed triple weibull distribution model is newly builed based on the large sample date,and reliability is evaluated by this model.Finally,the reliability evaluationresults of neural network model and mixed distribution model are compared.