Research On Thinning Algorithm For Massive Data And Its Software Implementation

With the development of electronic technology, the sampling frequency of electronic testequipment became higher, when the test equipment was monitoring for long times, therewould produce several hundred or even thousands of massive test data. According to theproblem that when making signal processing and feature extraction for massive test data, thespeed is slow and the feature is difficult to extract. The article carried out in-depth researchon massive data thinning algorithm, proposed a new heterogeneous data thinning algorithmfor massive test data, and made system research on signal processing method for the thinneddata and the reconstructed data. On the basis of this, designed the data processing system onthe base of LabVIEW, completed the thinning algorithm software. The article mainly hasresearch results as following:First of all, made research on massive data’s uniform data thinning and heterogeneousdata thinning algorithm in theory. Firstly, gave the basis of the data uniformly thinning theory;secondly according to the heterogeneous data thinning, proposed one kind of thinningalgorithm based on the curvature change, and made a applicably modify for thecharacteristics of test data; At last the paper gives a thinned data reconstruction method.Secondly, made a system research on the data signal processing method for the thinnedand reconstructed data, mainly made time frequency domain signal analysis method oftheoretical research on the parameters measurement, correlation analysis, energy spectrum,power spectrum, amplitude spectrum, phase spectrum, short-time FFT change, Wegener-Willdistribution and wavelet transform, the research provided theoretical support for the access toinformation of magnanimity data’s characteristics.Finally, on the basis of theoretical research, according to the system design scheme,chose the modular design concept, designed the data processing system on the platform ofLabVIEW, the system included five modules as user login, data normalization, signal analysis, data storage and report printing, and verified the system’s modules by simulation, and usedthe acoustic positioning system in underwater acoustic signals to experimental analysis andverification for the system’s thinning algorithm and other function modules. The results showthat, the system is capable of mass testing data for effective thinning, and can accuratelyextract the signal characteristic information; basically met the needs of the commonengineering test data processing, the system’s scheme also had strong feasibility.