Research On Gearbox Fault Diagnosis Technology Based On Swarm Intelligence Optimization Algorithm

As an important part of mechanical equipment,the gearbox has a wide range of applications in mechanical manufacturing field.The complexity and the harshness of its working environment directly determine the operation of the entire equipment.If the gearbox has a fault during operation,it will not only bring certain losses to the production efficiency,but also cause hidden dangers or even harm to personal safety.Therefore,in terms of ensuring the normal operation of equipment and protecting the safety of life and property,timely monitoring and diagnosis of the operating state of the gearbox has important application value and significance.This paper takes the laboratory gearbox as the research object,based on the fault type and failure mechanism of the gearbox.The time domain characteristic vibration signal of the gearbox under normal working conditions and several simulated fault conditions is collected by HG8916 data acquisition instrument.The data of the characteristic signal is saved in.txt format.After denoising the vibration signal by improving the threshold denoising method,this paper first analyzes the gearbox fault from the frequency domain angle.Secondly,based on the swarm optimization algorithm,the gearbox fault is analyzed from the perspective of time domain.The swarm intelligence optimization algorithm is a new algorithm developed in recent years.The core idea is to combine some simple individuals into a group and solve the problem of the target through a series of mechanisms and functions such as cooperation,intersection and learning.The swarm intelligence optimization algorithm covers many types of intelligent optimization algorithms,such as: Shuffled Frog Leaping Algorithm(SFLA)and Artificial Fish Swarm Algorithm(AFSA).This paper mainly studies the application of SFLA and AFSA intelligent optimization algorithms in gearbox fault diagnosis.By calculating the eigenvalues under various working conditions,the eight eigenvalues are selected as the judgment indicators and normalized..Then the eigenvalue index is divided into two parts: the training set and the test set.Combining SFLA and AFSA with Support Vector Machine(SVM)respectively,the SFLA-SVM and AFSA-SVM algorithm models are constructed for gearbox fault diagnosis.The final result is compared with the fault diagnosis results of PSO-SVM.By comparison,the following conclusions can be drawn: SFLA-SVM algorithm model shows lower fault diagnosis accuracy and stability.Therefore,the application of this algorithm in the field of gearbox fault diagnosis needs further discussion and research..AFSA has high efficiency global optimization ability and high accuracy.The classification ability of SVM is optimized,the accuracy of classification is improved and the SVM algorithm is prevented from falling into local optimum.The results show that AFSA-SVM shows good fault diagnosis and recognition ability in gearbox fault diagnosis,showing high accuracy and stability.