Research On Blade Tip Clearance Signal Acquisition & Real-Time Processing Technologies Based On FPGA

The blade tip clearance parameters of aero-engines are directly related to the operating efficiency and safety of the entire aircraft.The measurement of blade tip clearance parameters abroad has entered the test program of aero-engines,but the domestic research on blade tip clearance measurement started late,and there are still many key problems that need to be solved urgently.Among them,the high-speed acquisition and real-time processing of the tip clearance signal are the basis of the tip clearance signal data analysis and application.In this thesis,researches were carried out from the aspects of tip clearance signal processing method,system design,algorithm real-time processing,etc.A tip clearance signal processing method based on wavelet matching algorithm was proposed,a tip clearance signal acquisition and realtime processing system was developed,and experimental verification was carried out.The mainly work is as follows:1.The characteristics of the tip clearance signal and the need for tip clearance signal processing were analyzed.The overall plan of 12-channel high-speed acquisition and real-time processing of blade tip clearance signal was designed,and the key technologies of tip clearance extraction based on wavelet matching algorithm,real-time processing of wavelet matching algorithm based on FPGA,and multichannel synchronous acquisition and high-speed transmission were pointed out.2.A method of blade tip clearance signal processing based on wavelet matching algorithm was proposed.According to the waveform characteristics of the Gauss-like peak of the blade tip clearance signal,the mathematical model of the wavelet matching algorithm was derived;According to the model,the scale estimation and single-peak region segmentation method based on adaptive double thresholds,the peak position fitting method based on cubic matching and the peak height fitting method based on quintic wavelet matching were designed;The simulation results showed that the relative estimation error of the peak height of the proposed algorithm was only 0.3% under 10 dB signal-to-noise ratio,and the relative estimation error of the peak height within the speed range of 1000rpm-20000 rpm was less than 0.2%.3.A high-speed acquisition and real-time processing system for blade tip clearance signals was developed.A synchronous acquisition module for multi-channel signals was designed;A semi-parallel,multi-scale time-sharing matching hardware architecture based on FPGA was designed,and the transplantation of wavelet matching algorithm was realized.The simulation showed that the relative estimation error of the peak height of the transplanted algorithm in the speed range of 1000rpm-20000 rpm was less than 0.2%;The PCIe2.0 protocol was used to design a high-speed data transmission module;The upper computer software was designed based on Qt,which realized the functions of on-line monitoring and off-line analysis of blade tip clearance signals.4.An experimental platform for capacitive blade tip clearance measurement was built,and acquisition experiments and tip clearance measurement experiments were carried out.Experimental results showed that under offline processing,the maximum measurement error of the wavelet matching algorithm was 9μm in the speed range of1000rpm-4000 rpm,which was about 67% and 41% lower than the measurement errors of the direct peak finding method and the Caruana algorithm respectively;Under real-time processing,the maximum measurement error of the wavelet matching algorithm in the speed range of 1000rpm-4000 rpm was 11μm.