Research On The Correlation Relationship And Abnormal State Discrimination Of Sensor Data Of Special Vehicle Integrated Transmission Devic

Special vehicle equipment,such as tanks armored vehicles,infantry fighting vehicles and self-propelled artillery,is an important guarantee for the modernization of the army and the national security strategy.As the core component to realize the power transmission of special vehicle equipment and ensure its mobility performance,the power-shift steering transmission device will affect the mobility performance of special vehicle and cause serious economic and military losses once its running state is abnormal.Therefore,a variety of sensors are needed to monitor the power-shift steering transmission device for a long time to ensure that the special vehicle equipment can run safely and stably for a long time.This paper takes the power-shift steering transmission device of special vehicle as the research object,aiming at the problems that it is difficult to effectively distinguish the data changes caused by the working condition changes of the power-shift steering transmission device from the abnormal data caused by the device fault during the operation process,and that the sensor false alarm affects the driver’s judgment of the real operating state of the device,the following researches are carried out:First of all,in the face of the complex and changeable coupling relationship between the physical parameters of the power-shift steering transmission device,the sensor data association network was established,the sensor nodes with independent association relationship were analyzed.A single working condition was divided according to the power transmission mode and steering state of the device,and the strength change of the vehicle-mounted sensor data association relationship at different gears was studied.Aiming at the sensor data with complex changes,the stability of vehicle sensor data association under complex working conditions is studied.Secondly,density based spatial clustering of applications with noise(DBSCAN)was used to cluster the status monitoring data,so as to eliminate the interference of non-associated data on the accuracy of data reconstruction.The reconstruction model of state monitoring data was constructed by deep denoise autoencoder(DDAE)to obtain the deviation characteristics sensitive to abnormal data.The support vector data description(SVDD)algorithm was used to construct a hypersphere with the deviation characteristics of normal state monitoring data,so as to realize abnormal detection of comprehensive transmission device status monitoring data under complex working conditions.Finally,based on the analysis results of sensing data correlation,the correlation diagram of monitoring node of power-shift steering transmission device is drawn according to graph theory.A key node state prediction model based on long short-term memory(LSTM)framework is constructed.It is proved that the proposed method can accurately predict the state parameters of key nodes and increase the reliability of sensor data.By analyzing the correlation between the model prediction parameters and the associated sensor data,the abnormal state of the sensor data of the power-shift steering transmission device is judged,which provides a new abnormal state discrimination method for the condition monitoring and health management of the power-shift steering transmission device.