Fault Diagnosis Of Mechanical Transmission System Based On Adaptive Signal Processing

The prevention of artillery malfunction in the course of combat has always been the most important thing to prevent safety accidents.It is of important significance for artillery to carry out available research on fault diagnosis to accomplish high-intensity combat tasks in a short time.In the course of battle,the force of artillery transmission mechanism changes enormously,and its failure rate increases constantly.Therefore,this paper applies adaptive signal processing method to fault diagnosis of the transmission mechanism of artillery steering machine and high-altitude machine.Its main research contents are as follows:1)A fault diagnosis method of artillery steering machine based on LMD and permutation entropyIn the course of battle,the working environment of artillery steering machine is very wreched.So,the collected signals is often mixed with a large amount of noise.To solve this problem,a fault diagnosis method based on LMD and permutation entropy is proposed in this section.In this section,the validity of the method is verified by the gear fault data of Qianpeng Company,and then applied to the data of artillery steering machine.The diagnosis process is as follows:Firstly,the vibration signals of gear normal state,pitting fault and wear fault are decomposed by LMD algorithm to get each PF component;Then,calculate the permutation entropy of each PF component,and the best parameters of permutation entropy are verified by data,then the feature vectors with good resolution are obtained.Finally,support vector machine is used to complete classification.In order to prove the superiority of LMD algorithm,fault diagnosis method based on EMD and arrangement entropy is compared with the above method.From the point of view of diagnosis accuracy,the method adopted in this section obtains better results and has certain engineering significance.2)A fault diagnosis method of artillery elevating machine based on improved CEEMDAN and multiscale entropyIn order to solve the problem of non-stationary,non-linear and feature extraction difficulties of the vibration signal of elevating machine,this section a fault diagnosis method based on improved complete ensemble empirical mode decomposition with adaptive noise(CEEMDAN)and multiscale entropy is proposed.In this section,bearing fault data of Case Western Reserve University are used to verify the effectiveness of the method,and then applied to the data of artillery elevating machine.The diagnosis process is as follows:Firstly,a set of intrinsic mode functions(IMFs)are obtained by signal processing with improved CEEMDAN;then,the first six IMFs with the most fault information are selected,and fault features are extracted by multi-scale entropy(MSE);finally,the extracted eigenvectors are input into probabilistic neural network(PNN)to realize fault Diagnosis.In order to prove the superiority of the improved CEEMDAN-MSE method,it is compared with EMD-MSE,EEMD-MSE,CEEMDAN-MSE,improved CEEMDAN-AE and improved CEEMDAN-SE.The experimental results show that the method greatly improves the diagnostic accuracy and highlights the advantages of the method.3)GUI Design for Fault Diagnosis of Transmission SystemOn the basis of the previous two parts and with the help of GUIDE of MATLAB,"Gear Fault Diagnosis System Based on LMD and Permutation Entropy" and "Bearing Fault Diagnosis System Based on Improved CEEMDAN and Multiscale Entropy" are designed,which makes the diagnosis process convenient and fast,and also makes the process of diagnosis more integrated and practical.