Intelligent Diagnosis Method Of Planetary Gear Based On 1-DCNN Neural Network Of AdaBoost

Planetary gear mechanism is an indispensable part of most current mechanical equipment.However,because these construction machinery and equipment tend to have heavy workloads,poor operating environments,and unstable strong impact loads.Therefore,it is very important to study the rolling bearing and gear fault diagnosis methods of planetary gearboxes,which can ensure the safe operation of the machine and reduce economic losses.In the past fault diagnosis methods,there have been some problems.First,the collected data needs to be pre-processed,and it occupies a large part of the fault diagnosis workload;second,the existing traditional fault diagnosis methods have low diagnosis efficiency,and the intelligent diagnosis methods used such as convolution Neural networks also have problems such as insufficient stability after diagnosis,slow convergence speed,and easy overfitting.In this article,aiming at the existing problems,by discussing the research background of the subject,analyzing the research status and significance at home and abroad,through the expression of the existing methods,the common difficulties and points in the diagnosis of planetary gearboxes Perform analysis.Through the study of the failure modes of gears and bearings,the subsequent fault diagnosis test plan is planned for the use of the collected vibration signals of the faulty gears and bearings.Aiming at the problem that the vibration characteristics of planetary gearbox faults require preprocessing,identification difficulties,and slower convergence of the diagnostic model,1-DCNN is used to extract the features,which is characterized by the ability to directly learn features from the original time-domain vibration signal to complete the fault diagnosis.This method has two advantages;on the one hand,it can reduce the workload of data preprocessing;on the other hand,it can reduce the loss of characteristics of the original vibration signal during the processing.At the same time,the integrated learning theory is introduced,and the AdaBoost integrated learning algorithm is selected as the theoretical basis;the 1-DCNN fault diagnosis model based on AdaBoost is proposed as the planetary gearbox intelligent diagnosis method,which has more stable diagnosis output and can speed up the network model Convergence speed to prevent local optimization of the network and easy over-fitting.First,use 1-DCNN to extract features from the original time-domain vibration signal of the gear.Secondly,two classifiers with weaker classification performance are used.The weight update rule is based on the error rate of the classifier.After the weight update is completed,the classifier with weaker classification performance will continue the current training process.The whole process above will be repeated.Finally,the weak classifiers are integrated by setting the strategy of the integrated convolutional neural network.Establish a stable intelligent fault diagnosis model of planetary gear mechanism.In addition,the corresponding experiments are designed to verify the effectiveness of the algorithm and model.Experimental results show that the average recognition accuracy of the combined convolutional neural network fault diagnosis model reaches 98%,and the highest recognition accuracy reaches 99%.It can be seen from the experimental results that the 1-DCNN neural network based on AdaBoost proposed in this paper can quickly diagnose the original vibration signals of planetary gears.Compared with the traditional convolutional neural network for the diagnosis of the gear’s original time-domain vibration fault signal,the model in this paper has less workload;and at the same time has a more powerful fault diagnosis capability;the convergence speed is also further improved.The AdaBoost-based 1-DCNN neural network of the planetary gearbox established in this paper can diagnose the fault status of gears and bearings more efficiently;at the same time,it avoids the tedious and time-consuming signal preprocessing process in common fault diagnosis methods,and reduces the diagnosis.Time,and the recognition accuracy is greatly improved,which effectively improves the fault diagnosis efficiency of the planetary gearbox.Has certain practical application value.