福田 盛介

In this paper, a signal-processing method for a lunar lander using deep learning is proposed. The ability for pinpoint soft landing on a lunar/planetary surface broadens the range of scientific and exploration missions. To perform pinpoint landing, measurement of the relative velocity with respect to the surface is essential. Landing radar is a sensor that measures the relative velocity. To measure the velocity, the landing radar irradiates the surface with a pulse wave and observes the Doppler shift. High-precision measurement on complex terrains, a crater, or a slope has always been the problem of landing radar because the irradiated terrains strongly affect the accuracy. We propose a measurement system that performs with high accuracy on complex terrains using convolutional neural networks. Moreover, we confirm that the proposed method could improve the measurement accuracy compared with the existing method.

The exploration of energization and radiation in geospace (ERG) satellite, nicknamed "Arase," is the second satellite in a series of small scientific satellites created by the Institute of Space and Astronautical Science of the Japan Aerospace Exploration Agency. It was launched on December 20, 2016, by the Epsilon launch vehicle. The purpose of the ERG project is to investigate how high-energy (over MeV) electrons in the radiation belts surrounding Earth are generated and lost by monitoring the interactions between plasma waves and electrically charged particles. To measure these physical processes in situ, the ERG satellite traverses the heart of the radiation belts. The orbit of the ERG is highly elliptical and varies due to the perturbation force: the apogee altitude is approximately 32,200-32,300 km, and the perigee altitude is 340-440 km. In this study, we introduce the scientific background for this project and four major challenges that need to be addressed to effectively carry out this scientific mission with a small satellite: (1) dealing with harsh environmental conditions in orbit and electromagnetic compatibility issues, (2) spin attitude stabilization and avoiding excitation of the libration by flexible structures, (3) attaining an appropriate balance between the mission requirements and the limited resources of the small satellite, and (4) the adaptation and use of a flexible standardized bus. In this context, we describe the development process and the flight operations for the satellite, which is currently working as designed and obtaining excellent data in its mission.

<p>SLIM (Smart Lander for Investigating Moon) is the Lunar Landing Demonstrator which is under development at ISAS/JAXA. SLIM demonstrates not only so-called Pin-Point Landing Technique to the lunar surface, but also demonstrates the design to make the explorer small and lightweight. Realizing the compact explorer is one of the key points to achieve the frequent lunar and planetary explorations. This paper summarizes the preliminary system design of SLIM, especially the way to reduce the size.</p>

For planetary exploration missions using a lander, autonomous pinpoint landing capability whose precision is less than 100-meter must be needed in order to land on limited investigation areas. This capability cannot be realized with an inertial navigation system in terms of accuracy, so that the inertial error should be reduced in some way. One of solutions is image-based velocity measurement in a navigation path of the planetary lander. However, images taken by the lander have probrems such as frame rate limit and motion blur. Moreover, the available onboard resources of computation are limited. This paper proposes a velocity estimation method using a single blurred image for the pinpoint planatary landing. We present the method based on cepstral analysis and discuss feasibility through resource evaluation with a space-grade FPGA.

In recent years, investigating moon is once again drawing attentions of several countries. In Japan, an image-based autonomous navigation method is studied for a moon investigating project by the small lunar lander. The purpose of this paper is to propose a safe landing areas detection method for lunar environmental suitability that uses brightness values and to propose a landing point determination algorithm in a limited calculating area for the actual lunar lander. In addition, we study acceleration of the processing speed for hardware implementation and robustness of the proposed method.

The AC impedance of the LiFePO4 type lithium-ion secondary cells was initially measured, and charge/discharge cycling was constantly repeated at 10, 23, and 45 degrees C to understand their electrochemical characteristics. The decrease in the impedance for the charge transfer was observed at each temperature even though the SOC was set in 50%. We assumed from the impedance changing and discharge trend that the negative electrode performed as a "lithium-ion reservoir" by cycles. (C) The Electrochemical Society of Japan, All rights reserved.

<p>In this study, a crater detection method for a moon-landing system with low computational resources is proposed. The proposed method is applied to the Smart Lander for Investigating Moon (SLIM), which aims for a pin-point landing on the moon. According to this plan, surface images of the moon will be captured by a camera mounted on the space probe, and the craters are to be detected from the images. Based on the positional relationship between detected craters, the method estimates the exact flight position of the space probe. Because the computational resources of SLIM are limited, rapid and accurate crater detection must be performed using fixed-point arithmetic on a field-programmable gate array (FPGA). This study proposes a crater detection method that uses principal component analysis (PCA). The computational processing for crater detection by PCA is performed by product-sum operations, which are suitable for fixed-point arithmetic. Moreover, this method is capable of parallel processing; hence high-speed processing is expected. This study not only introduces a crater detection method using PCA but also evaluates the properties of this method.</p>

Lithium-ion secondary cells are widely used for the space applications, today. Among these applications, REIMEI, which was launched in 2005, was one of the first satellites using lithium-ion battery. The off-the-shelf type cells designed using spinel manganese oxide for the positive and the graphitized carbon for the negative electrode were used. The cell case was made of aluminum laminated film and the structure was reinforced by the aluminum case filled with epoxy resin. Today, ten years has passed, and the battery experienced 55,000 cycles for charge and discharge. The current distribution between two batteries almost coincided together even after the long term operation, which revealed the stable performance of the lithium-ion secondary cells under the microgravity in space. (C) The Electrochemical Society of Japan, All rights reserved.

A prototype lithium-ion battery with a bis(fluorosulfonyl)imide (FSI)-based ionic liquid electrolyte was developed. The prototype was mounted on a demonstration module of the "Hodoyoshi-3" microsatellite, which was successfully launched on June 20, 2014. Qualification tests for space application, including radiation tolerance and vacuum tests, revealed negligible degradation of the ionic liquid-based lithium-ion battery (IL-LIB) cell. According to the flight data, the IL-LIB cell can exist stably in an ultra-high vacuum environment despite its thin and flexible pouch casing without any rigid anti-vacuum reinforcements. Furthermore, the power unit showed the same charge-discharge performance as that predicted by the charge-discharge behavior of an identical cell on the ground, suggesting that the IL-LIB cell maintains performance in high vacuum a microgravity environment. These results prove that LIB cells with FSI-based ionic liquids can be used as a power source for space applications. (C) The Electrochemical Society of Japan, All rights reserved.

© 2014, The Author(s). HISAKI (SPRINT-A) satellite is an earth-orbiting Extreme UltraViolet (EUV) spectroscopic mission and launched on 14 Sep. 2013 by the launch vehicle Epsilon-1. Extreme ultraviolet spectroscope (EXCEED) onboard the satellite will investigate plasma dynamics in Jupiter’s inner magnetosphere and atmospheric escape from Venus and Mars. EUV spectroscopy is useful to measure electron density and temperature and ion composition in plasma environment. EXCEED also has an advantage to measure spatial distribution of plasmas around the planets. To measure radial plasma distribution in the Jovian inner magnetosphere and plasma emissions from ionosphere, exosphere and tail separately (for Venus and Mars), the pointing accuracy of the spectroscope should be smaller than spatial structures of interest (20 arc-seconds). For satellites in the low earth orbit (LEO), the pointing displacement is generally caused by change of alignment between the satellite bus module and the telescope due to the changing thermal inputs from the Sun and Earth. The HISAKI satellite is designed to compensate the displacement by tracking the target with using a Field-Of-View (FOV) guiding camera. Initial checkout of the attitude control for the EXCEED observation shows that pointing accuracy kept within 2 arc-seconds in a case of “track mode” which is used for Jupiter observation. For observations of Mercury, Venus, Mars, and Saturn, the entire disk will be guided inside slit to observe plasma around the planets. Since the FOV camera does not capture the disk in this case, the satellite uses a star tracker (STT) to hold the attitude (“hold mode”). Pointing accuracy during this mode has been 20–25 arc-seconds. It has been confirmed that the attitude control works well as designed.

We are in progress to develop a system for automatic operation of a satellite in order to reduce human load at satellite steady operation phase. The ground station for small satellite REIMEI (INDEX : INnovative-technology Demonstration EXperiment) is used as a test bench for verification of the proposed method. In our new automatic operation system, a scheduler software as a substitutive operator manages all the operations through a unified procedure, including sending command, receiving telemetry, and driving antenna in accordance with an operation time line which is prepared before the operation pass. The scheduler also performs diagnostics of satellite anomaly based upon the received telemetry data and status of the ground station. In case that some anomaly of the satellite is detected, the scheduler initiates an emergency schedule that was prepared depending on the emergency level. The automatic operation system is nearly completed for downlink operations of the data recorder that account for 75% of REIMEI steady operation. This approach is very effective to reduce psychological and physical load of operators.

本論文は,高分解能な合成開口レーダ(Synthetic Aperture Radar : SAR)画像のスペックル低減処理において,テクスチャやエッジなどの特徴量を保存する手法を提案している.提案する手法は,Gamma-Gamma MAPフィルタにおける局所領域からのオーダパラメータの推定に際して,オーダパラメータの推定精度を向上させるために,比較的広い領域を選択し,注目画素と同様の性質を有する領域に含まれる画素で構成されるウィンドウを新たに設定し,得られたウィンドウ内に含まれる画素を用いて,オーダパラメータを推定する.得られたオーダパラメータを基に,事後最大確率推定を行うことで,同一領域におけるテクスチャ情報を保持しつつ,スペックルを低減する.計算機シミュレーション画像による性能評価により,特徴量の保持性能と平滑性能を確認した上で,実際のSAR画像に適用することにより,従来手法と比較して高い視覚的自然性を確認した.

This paper describes the outline and the five years&apos; on-orbit results of the small scientific satellite REIMEI for aurora observations and demonstrations of advanced small satellite technologies. REIMEI is a small satellite with 72 kg mass, and is provided with three-axis attitude control capabilities for aurora observations. REIMEI was launched into a nearly sun synchronous polar orbit on Aug. 23rd, 2005, from Baikonur, Kazakhstan, by Dnepr rocket. REIMEI satellite has been satisfactorily working on the orbit for five years at present as of January, 2011. Three-axis control is achieved with accuracy of 0.1 degrees (3 sigma). Multi-spectrum images of aurora are taken with 8 Hz rate and 2 km spatial resolution to investigate the aurora physics. REIMEI is performing the simultaneous observation of aurora images and particle measurements. REIMEI indicates that even a small satellite launched as a piggy-back can successfully perform unique scientific mission purposes. (C) 2011 Elsevier Ltd. All rights reserved.

The Institute of Space and Astronautical Science JAXA is developing a landing radar comprising a radio altimeter and a velocity meter, which are two of the mandatory navigation sensors for a planetary lander. A BBM of the landing radar has been evaluated on natural terrains by using a helicopter. This paper introduces the BBM hardware and discusses the dynamic performance in field experiments.

Intensity fluctuation of synthetic aperture radar (SAR) images, due to speckle and texture, statistically brings series connections of dark pixels. Resolution enhancement of SAR systems can provide texture with correlation properties. In this paper, simulation studies are carried out to investigate the relation between the occurrence of the stochastic dark line features and texture correlation.

Support vector machines (SVMs), newly introduced in the 1990s, are promising approach to pattern recognition. They are able to handle linearly nonseparable problems without difficulty, by combining the maximal margin strategy with the kernel method. This paper addresses a novel SVM-based classification scheme of land cover from polarimetric synthetic aperture radar (SAR) data. Polarimetric observations can reveal existing different scattering mechanisms. As the input into SVMs, the polarimetric feature vectors, composed of intensity of each channel, sometimes complex correlation coefficients and textural information, are prepared. Classification experiments with real polarimetric SAR images are satisfactory, Some important properties of SVMs, for example the relation between the number of support vectors and classification accuracy, are also investigated.

Speckle statistically brings series connections of dark pixels, which can be observed as dark line features in synthetic aperture radar (SAR) images. The dark lines have no physical meaning. In this paper. line features of that kind in high-resolution SAR images whose intensity obeys a K-distribution are studied. It is stochastically explained that the dark line features in 1-look K-distributed images can be observed more distinctly than those in exponential distributed images. It is further revealed that such line features are detectable enough, even if the K-distributed images are multilooked. The experiments on simulated images as well as on actual SAR images confirm the explanation.

Texture is an essential key to the classification of land cover in SAR images. A wavelet-based texture feature set is derived in this paper. It consists of the energy of subimages obtained by the overcomplete wavelet decomposition of local areas in SAR images, where the downsampling between wavelet levels is omitted. The feature set has been successfully applied to multifrequency polarimetric images of the Flevoland site, an agricultural area in The Netherlands. The methods of polarization selection and feature reduction are also discussed.

The smoothing effect of the wavelet-based speckle filtering that we proposed is investigated. The filtering reduces the amplitude of wavelet coefficients, and a theoretical investigation with the Haar basis derives a functional relation between the ENL and two parameters: the wavelet level and the degree of the amplitude reduction.

A filter for suppressing speckle in synthetic aperture radar (SAR) images utilizing wavelet is proposed. The filter suppresses speckle by reducing the amplitude of the detail images in wavelet subspaces, while preserving edges by releasing the amplitude reduction around edges; information on edges, contained in the detail images, is utilized for edge detection. Simulations and application to SAR images have shown that the performance of the filter is satisfactory in both smoothing and edge preservation, and in generating visually-natural images as well.