Study And Analysis On Nonuniformly Sample Thermos In Modern Test Technology

With the development of scientific instrument and technical, digital instruments has been used in more and more fields, it has been attached more importance to signal sampling and signal processing. The nonuniformly sampled signal is an importance research component of digital signal sampling and processing subject, how can afford accurate sampled time base, or how can calibrate the sampled time base, and reconstruct the sampled signal sequence. Aimed on this research theme, this thesis studies some key issue of nonuniformly signal sampling theory and its application. the main contributions of this thesis are given as follows:1. In the case of determinacy time-base distortion, Author analysis two time-base distortion estimate arithmetic in common use. Because the Analytic Signalestimate arithmetic has its essence limitation, and the apply extension is poor. An unproved three points approached algorithm of time-base distortion estimation based on sine-fit algorithm is proposed. And the experiment proved it can advanced the estimation performance effectively.2. A new method based on windowing sinewave parameter estimation and interative algorithm is presented to reconstruct the nonuniformly sampled periodic signal. In this method a windowing parameter estimation algorithm is applied to obrain the initiative value of iteration for the nonuniformly sampled record. After that, we can get adjustment coefficient of the iterative algorithm based on the minimization of error energy of each step. Based on this, we can reconstruct the original signal. The simulation result shows the availability and superiority of the algorithm compared to exlisting algorithms.3. In the case of uncertainly part timing jitter, we present an analysis of the effects of timing jitter of the waveform samples. Since only the nominal sample times are know, the relation of the timing jitter distribution probability density function and the signal frequency spectrum.4. with the development of real time signal processing, recently, on many real time sampled system, adopted forepart parallel sampling , backpart multi-serial processing technique. The author argument the uniform sampled model and the frequency spectrum, and investigate the sampled periods signal reconstruction in this Model.