The Statistic Law And Fractal Characteristic Of Aerosol's Scattering Pulse Signal Counting Distribution

In this paper, founding on the basic concept of aerosol mass subset, the statistical law of the scattering light pulse signal counting distribution is studied, by using two-dimension parameters analytical method. The basic parameters distributing functions'comparability of the pulse signal is researched. Based on which the power function relation between the two statistical parameters (amplitude and width) is investigated.First, the significance of recording the counting distribution based on the observed signal's measurement is expounded. And then, it is proved that any significant direct observation results all can be expressed as a counting sequentially statistical format on the basis of natural number, be based upon starting from the measurement process of commonly sequential pulse signal and combining the typical probability distribution function. Besides, the random pulse signal aggregate element subset's self-similarity and statistically fractal concept's measuring significance are analyzed.For the purpose of improving the counting efficiency and sampling flux, the light scattering counting sensor's basic configuration is investigated. Based on which, two 28.3L/min flux photoelectrical sensors are designed, accompanied with the protecting air casing. By using the two high sensitivity photoelectrical sensors, the scattering random pulse signal's amplitude V and widthτdistribution regulations of different standard particles and aerosols are investigated. The results indicate that the two functions are all have logarithmical normal distribution formats. In further, the metrical results of photoelectrical sensors'pulse signal amplitude distribution of the background noise is analyzed and studied. Based on which it is found that the relation between the statistical distribution and the signal's random characteristic also agrees the logarithmical normal distribution. Besides, it is also proved that the distributing statistical parameterμιn=lnιhas relation with the precision of the measuring system, whileσιn=(?) reflects the signal's inner structure and the intensity of the measuring process's uncertain factor.Random counting pulse signal distribution function's basic character is studied theoretically. Integrating the limited basic characteristic of pulse signal's measuring data range,the relation between the counting pulse signal distribution function's separate variable format,normalized factor and the measuring data range is deduced,and the influence of two statistic parametersμιn. andσιn on the distribution function's geometric monotony is discussed. And then, the comparability of the distribution function in nonlinear convergent-divergent transformation format is researched.Besides, the comparability of the amplitude and width distribution functions is also analyzed. Finally,it is found that the amplitude and width meet statistically fractal relation who can be desribed by ((τ-τm)/τ)∝((ν-νm)/ν)α,while the fractal dimensionαis defined asα=σιnν/σιnτ.