Function Test Platform And Evaluation Method Research Of Typical Mechanism Of A Launcher

As an important subsystem of the gatling weapon system,the launcher has the characteristics of complex internal structure,poor working conditions and frequent working times.Under the action of continuous and high-frequency complex impact loads,it is prone to produce complex dynamic response problems such as vibration,shock,deformation,temperature rise and flutter.Moreover,the fault formation location of launcher is very hidden and the specific fault type is difficult to distinguish,which brings great difficulties to the realtime detection,evaluation and diagnosis of weapon system.Taking a certain type of gatling weapon as the research object,aiming at the problems of high failure rate,great harm and difficult to identify fault location,this paper analyzes the typical mechanism with high failure rate in the launcher.The monitoring and diagnosis of the typical mechanisms of the launcher can predict the failure in advance,diagnose the damage in time and test accurately,so as to ensure the reliability of the launcher.Based on the analysis of structure composition,working process and working principle of the launcher,the cause of failure of typical mechanisms and corresponding failure influence parameters were obtained,and the specific failure type and failure incidence rate of the mechanisms were determined by fault statistical analysis.The sensor is used to collect the fault influence parameters,and all the parameters are classified according to the influence parameters corresponding to the typical mechanisms,and the data is used as the main basis of the fault detection platform,which mainly includes three modules:fault data processing,online fault monitoring and intelligent fault diagnosis.First of all,for the fault data processing module of typical mechanisms,there are null values,zero values,maximum values and some impurity noises in the data collected by the sensors.Data processing to obtain more accurate fault online monitoring and fault intelligent diagnosis test data.Secondly,for the faults online monitoring module of typical mechanisms,using HilbertHuang Transform method,the EMD decomposition diagram of the input signal,the energy distribution diagram of IMF component,the HHT spectrum and the HHT side spectrum are obtained,and the change characteristics of the signal in the time domain,frequency domain and amplitude domain are displayed by the graphical way.By comparing the results of HHT results of normal parameters and fault parameters,the general fault characteristics are obtained,and the fault prediction is based on this,so as to achieve real-time online monitoring.Finally,for the typical mechanisms fault intelligent diagnosis module,the basic principle and diagnosis process of BP neural network and PSO-BP neural network are studied.The iteration times,running time,test accuracy and diagnosis results of the two algorithms are compared respectively,which can be found that the PSO algorithm has the characteristics of easy implementation,less iteration times and high detection accuracy.Moreover,compared with BP neural network,the running time is longer,but the final results are the same by the two neural networks,which realizes the function of fault diagnosis.