Testing And Data Preprocessing Of Dynamic Torque For Tracked Vehicle Transmission System

Load spectrum is the basis for fatigue life and reliability design of drive system components. The testing effectiveness of the dynamic load directly determines whether the compiling of load spectrum is reliable or not. It is the basis for further analysis and statistics of load data through the analysis of the effectiveness of dynamic torque signal and data preprocessing. There is no integrated load spectrum data from typical pavement test for reliability design and evaluation of tracked vehicle transmission system used in domestic. This thesis is aimed at the research that how to provide effective dynamic torque data for compiling dynamic load spectrum of tracked vehicle transmission system. It mainly studied the dynamic torque testing of tracked vehicle transmission system, analysis of testing accuracy of dynamic torque, data preprocessing and evaluation methods.In order to obtain the real dynamic torque of transmission system components, a dynamic torque testing system for tracked vehicle was established first. Aiming at some special requirements for tracked vehicle, the influence factors of testing were analyzed in detail and the function of the torque sensor suitable for the testing requirements was evaluated. The high speed synchronization acquisition of torque signal with various vehicle signals was realized and a variety of real vehicle testing on earth road and cement road with different operating condition was carried out.On the basis of the tracked vehicle dynamic torque testing, a method to analyze and evaluate testing accuracy is presented in this paper. The credibility of testing system was analyzed by probability statistics model and the frequency and amplitude of testing signal were respectively analyzed by power spectrum analysis and waveform analysis combining with the vehicle dynamic model. It ensures the macro effectiveness of testing data.In order to satisfy requirements of compiling dynamic load spectrum for tracked vehicle transmission system finally, the data preprocessing methods of how to deal with the irregular noise that usually exists in the dynamic load signal were studied. Boxplot method and standard variance test method were used to detect and remove spikes and finite impulse response filter with strict linear phase frequency characteristics was used to filter out high frequency interference signal. Then, the wavelet analysis used for dynamic load signal denoising was studied, and the calculation method of adaptive threshold developed from universal threshold was presented. The method solves the problem of how to remove noise of different frequency bands on the premise of keeping useful medium-high frequency dynamic load signal.At last, the results of different data preprocessing methods were evaluated based on theory of fatigue damage. The adaptive threshold de-noising method shows that it has a better effect on dynamic load signal de-noising. It also proves that the processed data conforms to the requirements of compiling dynamic load spectrum for tracked vehicle transmission system.