Statistics Law Of Light Scattering Signals And Its Application In Particles Counting

With the increase of hazy days, the public is paying more attention to the pollution of the micro-particles suspended in the air. Besides, with the development of science and technology, the requirements for cleanliness is becoming higher and more strict. Therefore, the study of suspended particles and its measuring instruments turns into an important research spot these days. In order to improve the laser scattering particle counter, this Ph.D thesis establishes an analysis model of light scattering particle random measuring process based on the analysis of the process of measuring the particle diameter by light scattering method. Combining with dual-parameter analysis of random pulse signal, this Ph.D thesis also studies the amount of statistical regularity and self-similarity which provides theoretical guidance for the design and optimization of the light scattering particle counter.Firstly, this Ph.D thesis studies optical field distribution of sensor with theoretical analysis and numerical simulation. With the research on kinds of independent random factors using dual-parameter analysis method according to the basic parameters as amplitude and width, analysis model of light scattering particle random measuring process is proposed. Moreover, the nonlinear constraint relation on statistical distribution functions in light scattering particles random measuring process is derived. The results show lognormal distribution function with natural numbers as the independent variables satisfies the nonlinear constraint relation.Then, in order to understand light scattering particle measuring system better, statistical distribution law of different measuring factors as standard particles, random air suspend particles, the background noise and different characteristics of background noise in the same light scattering particle measuring system are studied. The experimental results indicate that the amplitude and width distribution of light scattering random signal group, the distribution of signal subset and other statistical distribution of factors with different characteristics fit well with lognormal distribution function with natural numbers as the independent variables. Tested by chi-square distribution fitting and correlation coefficient, lognormal distribution is more accurate than normal distribution with a higher correlation coefficient of0.95and acceptable significant level of0.995. Besides, on the basis of studying statistical laws of different measuring factors, the nonlinear transformation relations between statistical distribution of different physical factors are deduced using statistical fractal theory which could simplify the calculation procedures of the fractal dimension of random particle. Finally, combining with theoretical analysis and experimental research, based on the design of light scattering particle measuring sensor, lognormal distribution is applied to the optimization of measuring channels with1/3channels cutoff which could expand the measuring range of the sensor. moreover it is also used in the optimization of particle size resolution, and particle size resolution has increased by9%compared to normal distribution. Meanwhile, the light scattering particle measuring sensor is improved in many aspects as lighting system, the photosensitive area, background noise, detection system and gas system, etc. Besides, the invention patent authorization of the sensor is obtained. The systematic parameters as counting efficiency particle size resolution of the light scattering particle device have reached the requirements of international standards at the large flow measuring mode in practical measurement. Moreover, the measuring sensor has the advantages of small size, low cost and stable performance.