| As a kind of complex atmospheric processes, the cloud (fog) plays an important role in the research of meteorological and atmospheric science. Cloud (fog) drops spectral distribution is an important parameter of cloud microstates, and research of drops spectral distribution helps to reveal the physical state of the atmosphere change process, Chemical processes, Mechanisms and laws occurred during the process, improve the ability of human predicting the atmosphere, preventing the ability of disaster weather and developing and utilizing the atmospheric resources. Therefore, based on the above, the particles detection system was proposed in this paper, in which wide-angle side light scattering system was designed, and scattering model simulation was conducted. In addition, the hardware system and software algorithm were designed for the testing system. The main researches of this paper are as follows:Firstly, several detecting methods of particles was introduced and it was determined to use the light scattering to measure the size distribution of particles in this paper. The Mie scattering theory was used to establish the scattering model of single particle, and the scattering characteristics of particle swarm optimization was analyzed by simulation software. The E-D model was established, and the effects of different scattering angles measured to the light of scattering energy measured were compared. Combined with the advantages of off-axis lighting in the airborne system, the different axis large angle lighting system was determined in this subject.Secondly, the bimodal unimodal distribution model was established by using the lognormal distribution, and the relationships between the light distribution model and the spatial distribution were derived by using the Mie scattering theory. Then the basic inversion algorithm was introduced, combined with the light energy coefficient matrix and the spatial distribution of light energy matrix inverse particle system, and compared the particle size distribution with the simulation.Then the principle of particle detection system and its the main component part were introduced, and make a note of the main part:Light source frame modulation, surface array CCD image sensor unit, principle of data collection, controlling and processing based on DSP chip and analysis and selection of related devices.Finally, the design of the hardware and software of the system was finished. Wherein, the hardware design is composed of power supply module circuit, analog video decoding circuit, SDRAM memory expansion circuit, laser driver control and so on. The DSP/BIOS operating system kernel was introduced in the software design, in which the software workflow of the system was illustrated.In this paper, the energy distribution characteristics of Mie scattering were obtained by Matlab simulation. The energy angle distribution model was established, and by comparing the data it was found that in the same case, the greater as the acquisition of light angle is, the greater the light energy E is, which make up for the lack of information collected by the small angle method. The size distribution in single-double peak particles and particles set were compared by using the direct and iterative methods to test the distribution of bimodal unimodal distribution particles. It was found that when the error parameter was used to analyze the data, the difference parameter from iterative algorithm is much less than that from direct method of inversion. A stable solution could be obtained by iteration of a certain number of times. And the iteration method is simple and with little constraint conditions, which will be helpful to transplant the subsequent DSP platform algorithm. |
PDF Full Text Request