Grant-in-Aid for Overseas Scientific Survey.
|Section||University-to-University Cooperative Research|
|Research Institution||Kobe University|
KITAMURA Shinzo Faculty of Engineering, Kobe University, Professor, 工学部, 教授 (80029131)
安藤 四一 神戸大学, 工学部, 助教授 (30031115)
賀谷 信幸 神戸大学, 工学部, 助教授 (30093503)
松本 治弥 神戸大学, 工学部, 教授 (80031053)
吉村 武晃 神戸大学, 工学部, 助教授 (70031127)
SIGELMANN R. ワシントン大学, 電気工学科, 教授
LEUNG TSANG ワシントン大学, 電気工学科, 教授
石丸 昭 ワシントン大学, 電気工学科, 教授
ANDO Yoichi Faculty of Engineering, Kobe University, Associate Professor
KAYA Nobuyuki Faculty of Engineering, Kobe University, Associate Professor
MATSUMOTO Haruya Faculty of Engineering, Kobe University, Professor
TSANG Leung Department of Electrical Engineering, University of Washington, Professor
YOSHIMURA Takeaki Faculty of Engineering, Kobe University, Associate Professor
ISHIMARU Akira Department of Electrical Engineering, University of Washington, Professor
SIGELMANN Rubens A. Department of Electrical Engineering, University of Washington, Professor
|Project Fiscal Year
1989 – 1990
Completed(Fiscal Year 1990)
|Budget Amount *help
¥3,900,000 (Direct Cost : ¥3,900,000)
Fiscal Year 1990 : ¥1,900,000 (Direct Cost : ¥1,900,000)
Fiscal Year 1989 : ¥2,000,000 (Direct Cost : ¥2,000,000)
|Keywords||Scattering / Absorption / Inverse problem / Neural network / Optical sensing / Ultrasound / Electro-magnetic wave / Laser radar / 散乱 / 吸収 / 逆問題 / ニュ-ラル・ネットワ-ク / 光 / 音波 / 電磁波 / レ-ザ-レ-ダ|
We have carried out a cooperative research with the University of Washington on multiple scatterings and inverse problems in 1989 and 1990. The major results are summarized in the followings.
1) Application of Neural Network to Inverse Scattering Problems.
The neural network technique has been applied to solve the inverse scattering problems. The fundamental idea is to first train the neural network by priori results. The network can speedily and efficiently solve the inverse problem, once it has been trained. We verified the validity of this technique by applying it to find particle-size distributions of aerosol from remote sensing data.
2) Multiple Scattering Effects.
We have made detailed studies of scattering by very rough surfaces by theoretical analyses, numerical simulations, and millimeter wave experiments. The results indicate that backscattering enhancement is sensitive to the roughness spectrum. The theoretical solutions provide a physical explanation of backscattering enhanceme
nt and agree well with the numerical results.
A new method has been proposed to measure the surface roughness by using multiple scattering effects. The two roughness parameters of rms heights and correlation lengths can be uniquely determined by twice measuring the second order moment of integrated intensities.
We have also examined a dense medium model consisting of two layers of snow on top of ice. Our results show that the copolarized cross sections are strong functions of frequency and snow-grain size. We have also found that a small number of the larger snow grains contribute significantly to the copolarized backscattering cross section.
3) Scattering of High Power Microwave by Ionospheric Plasma.
High power microwave may be scattered by the ionospheric plasma, which is transparent for the weak microwave. A rocket experiment was carried out for transparent for the weak microwave. A rocket experiment was carried out for studying the nonlinear effects of high power microwave. It showed that the electron cyclotron harmonic waves are excited by the microwave. Since the wave excitation mechanism is not well-understood, we have started to develop the computer simulation codes. Less