Publishing type：Research paper (scientific journal)   Publisher：American Meteorological Society  

Abstract

In this study, a clear-air turbulence event was reproduced using a high-resolution (250 m) large-eddy simulation in the Weather Research and Forecasting (WRF) Model, and the resulting wind field was used in a flight simulation to estimate the vertical acceleration changes experienced by an aircraft. Conditions were simulated for 16 December 2014 when many intense turbulence encounters (and one accident) associated with an extratropical cyclone were reported over the Tokyo area. Based on observations and the WRF simulation, the turbulence was attributed to shear-layer instability near the jet stream axis. Simulation results confirmed the existence of the instability, which led to horizontal vortices with an amplitude of vertical velocity from +20 to −12 m s⁻¹. The maximum eddy dissipation rate was estimated to be over 0.7, which suggested that the model reproduced turbulence conditions likely to cause strong shaking in large-size aircraft. A flight simulator based on aircraft equations of motion estimated vertical acceleration changes of +1.57 to +0.08 G on a Boeing 777-class aircraft. Although the simulated amplitudes of the vertical acceleration changes were smaller than those reported in the accident (+1.8 to −0.88 G), the model successfully reproduced aircraft motion using a combination of atmospheric and flight simulations.

DOI： 10.1175/jamc-d-21-0071.1

Scopus

Other Link： https://journals.ametsoc.org/downloadpdf/journals/apme/61/5/JAMC-D-21-0071.1.xml

Language：Japanese   Publisher：公益社団法人 日本生物工学会  

DOI： 10.34565/seibutsukogaku.100.4_189

Language：English   Publishing type：Research paper (scientific journal)   Publisher：Informa UK Limited  

ABSTRACT

Several industries require getting information of products as soon as possible during fermentation. However, the trade-off between sensing speed and data quantity presents challenges for forecasting fermentation product yields. In this study, we tried to develop AI models to forecast ethanol yields in yeast fermentation cultures, using cell morphological data. Our platform involves the quick acquisition of yeast morphological images using a nonstaining protocol, extraction of high-dimensional morphological data using image processing software, and forecasting of ethanol yields via supervised machine learning. We found that the neural network algorithm produced the best performance, which had a coefficient of determination of >0.9 even at 30 and 60 min in the future. The model was validated using test data collected using the CalMorph-PC(10) system, which enables rapid image acquisition within 10 min. AI-based forecasting of product yields based on cell morphology will facilitate the management and stable production of desired biocommodities.

DOI： 10.1093/bbb/zbab188

PubMed

Other Link： https://academic.oup.com/bbb/article-pdf/86/1/125/41881825/zbab188.pdf

J-GLOBAL

Publishing type：Research paper (scientific journal)   Publisher：Optica Publishing Group  

Herein, a calibration procedure to determine the depth positions of particles in a microfluidic channel via astigmatism particle tracking velocimetry (APTV) has been described. A neural network model focusing on the geometrical parameters of distorted particle images was developed to calibrate APTV. To demonstrate the efficiency of this procedure, the Poiseuille flow and depth of the particles, and dispersions in the microchannel were studied. The depth positions were determined with an uncertainty of . The present results suggest that the particle position dispersion could be a result of the degree of particle image deformation and its deviation.

DOI： 10.1364/ao.427571

Publisher：Copernicus GmbH  

Abstract. The control technique in a gust alleviation system by using the airborne Doppler Lidar technology is expected to enhance aviation safety to minimize the risks of turbulence-related accidents. Accurate measurement and estimation of the vertical wind velocity are very important in the successful implementation of a gust alleviation system by using the airborne Doppler Lidar technology. An estimation algorithm of airflow vector based on the airborne Lidars is proposed and investigated for preview control to prevent turbulence-induced aircraft accidents in flight. The use of the simple vector conversion method, which is an existing technique, assumes that the wind field between the Lidars is homogeneous. The assumption of a homogeneous field would be wrong when turbulence occurs due to large wind velocity fluctuation. The proposed algorithm stores the line-of-sight (LOS) wind data with each passing moment and uses recent and past LOS wind data in order to estimate the airflow vector. The recent and past LOS wind data are used to extrapolate the wind field between the airborne twin Lidars. Two numerical experiments – ideal vortex model and numerical weather prediction – were conducted to evaluate the estimation performance of the proposed method. The proposed method has much better performance than simple vector conversion in the two numerical experiments, and it can estimate accurate two-dimensional wind field distributions unlike simple vector conversion. The estimation performance and the computational cost of the proposed method can satisfy the performance demand for preview control.

DOI： 10.5194/amt-2020-106

Other Link： https://kaken.nii.ac.jp/grant/KAKENHI-PROJECT-18H03809/

Authorship：Lead author  

Language：English   Publishing type：Research paper (scientific journal)  

Classification of the biological activities of chemical substances is important for developing new medicines efficiently. Various machine learning methods are often employed to screen large libraries of compounds and predict the activities of new substances by training the molecular structure-activity relationships. One such method is graph classification, in which a molecular structure can be represented in terms of a labeled graph with nodes that correspond to atoms and edges that correspond to the bonds between these atoms. In a conventional graph definition, atomic symbols and bond orders are employed as node and edge labels, respectively. In this study, we developed new graph definitions using the assignment of atom and bond types in the force fields of molecular dynamics methods as node and edge labels, respectively. We found that these graph definitions improved the accuracies of activity classifications for chemical substances using graph kernels with support vector machines and deep neural networks. The higher accuracies obtained using our proposed definitions can enhance the development of the materials informatics using graph-based machine learning methods.

DOI： 10.1002/minf.201800155

PubMed

Language：Japanese   Publisher：土木学会  

CiNii Books

Other Link： https://kaken.nii.ac.jp/grant/KAKENHI-PROJECT-18K13834/

Publishing type：Part of collection (book)  

© Springer Nature Switzerland AG 2020. A new technique that integrates low dimensional model (LDM) based on proper orthogonal decomposition (POD) and the flight data of a commercial aircraft is proposed to realize real-time prediction of wind and atmospheric turbulence for aviation safety and efficiency. The proposed technique sequentially assimilates flight data into LDM and predicts the wind and atmospheric turbulence at lower computational cost than the general numerical weather prediction (NWP). Actual experiments were conducted for two cases: First, weather conditions of an extratropical cyclone approaching Japan, and second, stationary front in the sea near Japan. The actual experiments consisted of two cases: under the condition of an extra-tropical cyclone approaching Japan (Case 1) and a stationary front at Pacific Ocean near Japan (Case 2). In Case 1, the proposed method was able to produce matches between the areas predicted for turbulence and the locations where turbulence was actually encountered. The proposed method is able to correct these spatiotemporal uncertainties by using the flight data. In Case 2, NWP predicted weaker wind than the flight data, and the difference between the wind rates of the NWP and the flight data was about 10 ms−1 at 55 min after the take-off, which is the time of maximum wind magnitude by the flight data. The proposed method was able to correct this difference, and predict the maximum wind magnitude accurately.

DOI： 10.1007/978-3-030-27053-7_42

Scopus

Authorship：Lead author  

Publishing type：Research paper (international conference proceedings)  

© 2019 by German Aerospace Center (DLR). Published by the American Institute of Aeronautics and Astronautics, Inc. This paper explores a robust optimal guidance algorithm to determine an optimal cruise airspeed under weather uncertainty satisfying the required time of arrival at a specifying fix assigned by air traffic control. In order to alleviate the effect of weather prediction errors and improving the accuracy of time of arrival at the fix, the robust optimal guidance strategy incorporates a probabilistic weather prediction technique, which periodically up-dates the weather prediction data and estimates the weather prediction errors during flight by combining pre-inputted probabilistic weather prediction data and aircraft observation data. By using the updated probabilistic weather prediction data, the proposed robust optimal guidance strategy is performed to determine the optimal cruise airspeed meeting the required time of arrival with highest probability under the weather uncertainty. Through numerical simulations, the performance and effectiveness of the robust optimal guidance algorithm are evaluated and demonstrated.

DOI： 10.2514/6.2019-1639

Scopus

Publishing type：Research paper (scientific journal)  

© 2017 Royal Meteorological Society This study proposes an algorithm that combines ensemble numerical weather-prediction model data and aircraft flight data in a wind nowcasting system for safe and efficient aircraft operation. It uses an ensemble-weighted average method based on sequential importance sampling (SIS), which is a particle filter method for forecasting the wind field in real time. SIS is applied to the ensemble forecast data and control run data of the European Centre for Medium-Range Weather Forecasts (ECMWF), Japan Meteorological Agency (JMA), Korea Meteorological Administration (KMA), National Centers for Environmental Prediction (NCEP) and United Kingdom Met Office (UKMO) for the two case studies that use flight data from 72 commercial aircraft flights. The results show that SIS can forecast better than the other four methods: direct ensemble average (DEA), elite strategy (ES), and selective ensemble average (SEAV) and weighted average (SEWE), with average improvements in forecast performance of about 10–15%, even at 300 min ahead. In addition, the overall forecast performance between the forecast wind and observation of the radiosonde of SIS was slightly better than DEA. In both cases, the forecast performance was significantly improved on points along the flight path of the aircraft used for this study. Case analyses and the impact of differences in the hyper-parameters of SIS on forecast performance are also presented in this study.

DOI： 10.1002/met.1704

Web of Science

Scopus

Language：English   Publishing type：Research paper (scientific journal)   Publisher：TAYLOR & FRANCIS LTD  

© 2016 Informa UK Limited, trading as Taylor & Francis Group. An integrated method consisting of a proper orthogonal decomposition (POD)-based reduced-order model (ROM) and a particle filter (PF) is proposed for real-time prediction of an unsteady flow field. The proposed method is validated using identical twin experiments of an unsteady flow field around a circular cylinder for Reynolds numbers of 100 and 1000. In this study, a PF is employed (ROM-PF) to modify the temporal coefficient of the ROM based on observation data because the prediction capability of the ROM alone is limited due to the stability issue. The proposed method reproduces the unsteady flow field several orders faster than a reference numerical simulation based on Navier–Stokes equations. Furthermore, the effects of parameters, related to observation and simulation, on the prediction accuracy are studied. Most of the energy modes of the unsteady flow field are captured, and it is possible to stably predict the long-term evolution with ROM-PF.

DOI： 10.1080/10618562.2016.1198782

Web of Science

Scopus

Language：English   Publishing type：Research paper (scientific journal)   Publisher：AMER METEOROLOGICAL SOC  

© 2016 American Meteorological Society. The phased-array weather radar (PAWR) is a new-generation weather radar that can make a 100-m-resolution three-dimensional (3D) volume scan every 30 s for 100 vertical levels, producing ~100 times more data than the conventional parabolic-antenna radar with a volume scan typically made every 5 min for 15 scan levels. This study takes advantage of orders of magnitude more rapid and dense observations by PAWR and explores high-precision nowcasting of 3D evolution at 1-10-km scales up to several minutes, which are compared with conventional horizontal two-dimensional (2D) nowcasting typically at O(100) km scales up to 1-6 h. A new 3D precipitation extrapolation system was designed to enhance a conventional algorithm for dense and rapid PAWR volume scans. Experiments show that the 3D extrapolation successfully captured vertical motions of convective precipitation cores and outperformed 2D nowcasting with both simulated and real PAWR data.

DOI： 10.1175/WAF-D-15-0063.1

Web of Science

Scopus

Language：English   Publishing type：Research paper (scientific journal)   Publisher：IOP PUBLISHING LTD  

© 2015 The Japan Society of Fluid Mechanics and IOP Publishing Ltd. An integrated method of a proper orthogonal decomposition based reduced-order model (ROM) and data assimilation is proposed for the real-time prediction of an unsteady flow field. In this paper, a particle filter (PF) and an ensemble Kalman filter (EnKF) are compared for data assimilation and the difference in the predicted flow fields is evaluated focusing on the probability density function (PDF) of the model variables. The proposed method is demonstrated using identical twin experiments of an unsteady flow field around a circular cylinder at the Reynolds number of 1000. The PF and EnKF are employed to estimate temporal coefficients of the ROM based on the observed velocity components in the wake of the circular cylinder. The prediction accuracy of ROM-PF is significantly better than that of ROM-EnKF due to the flexibility of PF for representing a PDF compared to EnKF. Furthermore, the proposed method reproduces the unsteady flow field several orders faster than the reference numerical simulation based on the Navier-Stokes equations.

DOI： 10.1088/0169-5983/47/5/051403

Web of Science

Scopus

Publishing type：Research paper (international conference proceedings)  

© 2015, American Institute of Aeronautics and Astronautics Inc. All rights reserved. An integrated method consisting of a proper orthogonal decomposition (POD)-based reduced-order model (ROM) and a particle filter (PF) is proposed for real-time prediction of low-level turbulence. The proposed method is validated using identical twin experiments of the low-level turbulence case at Shonai airport in Japan. The results of the experiment show ROM with the estimated temporal coefficients by PF reproduces the wind velocity fluctuations similar to those produced by large eddy simulation with much lower computational cost. In addition, the time evolution of the wind field obtained by ROM-PF show that turbulent spatial structure is expressed accurately in comparison with the reference flow field. It is confirmed that the root mean square error of head and crosswind velocities decreased over time by repeating data assimilation.

Scopus

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

© 2014 International Society of Information Fusion. An integrated method of a proper orthogonal decomposition (POD) based reduced-order model (ROM) and data assimilation is proposed for real-time prediction of an unsteady flow field. In this paper, a particle filter (PF) and an ensemble Kalman filter (EnKF) are employed for data assimilation and the difference of predicted flow fields is evaluated in detail. The proposed method is validated using identical twin experiments of an unsteady flow field around a circular cylinder at Reynolds number of 1000. The PF and EnKF are employed to estimate coefficients of the ROM based on observed velocity components in the wake of the circular cylinder. The proposed method reproduces the unsteady flow field several orders faster than the reference numerical simulation based on Navier-Stokes equations. Furthermore, the prediction accuracy of ROM-PF is significantly better than that of ROM-EnKF. It is due to the flexibility of PF for representing a predictive probability density function compared to EnKF.

Web of Science

Scopus

Other Link： https://dblp.uni-trier.de/conf/fusion/2014

Publishing type：Research paper (scientific journal)   Publisher：Japan Society for Aeronautical and Space Sciences  

DOI： 10.2322/jjsass.61.159

Language：English   Publishing type：Research paper (international conference proceedings)   Publisher：IEEE  

The proper orthogonal decomposition (POD) and the particle filter (PF) are applied to estimate the low-level turbulence around an airport. POD is introduced to represent a wind velocity field by a reduced-order model. PF is employed to estimate coefficients of POD by using observed velocity components from the broadband radar located at an airport. In this study, a numerical experiment was conducted to investigate the effectiveness of the proposed method by comparing the estimated wind velocity field with the original wind velocity field obtained from large eddy simulation (LES). From the comparison results, the variation of turbulence intensity obtained from the present approach was found similar to the original LES result. The main advantage of the present approach is much lower computational cost compared to LES. © 2013 ISIF ( Intl Society of Information Fusi.

Web of Science

Scopus

Other Link： https://dblp.uni-trier.de/conf/fusion/2013

Announcement no：特開2019-179404 

Announcement no：特開2020-061007