Domain knowledge integration into deep learning for typhoon intensity classification

In this report, we propose a deep learning technique for high-accuracy estimation of the intensity class of a typhoon from a single satellite image, by incorporating meteorological domain knowledge. By using the Visual Geometric Group’s model, VGG-16, with images preprocessed with fisheye distortion, which enhances a typhoon’s eye, eyewall, and cloud distribution, we achieved much higher classification accuracy than that of a previous study, even with sequential-split validation. Through comparison of t-distributed stochastic neighbor embedding (t-SNE) plots for the feature maps of VGG with the original satellite images, we also verified that the fisheye preprocessing facilitated cluster formation, suggesting that our model could successfully extract image features related to the typhoon intensity class. Moreover, gradient-weighted class activation mapping (Grad-CAM) was applied to highlight the eye and the cloud distributions surrounding the eye, which are important regions for intensity classification; the results suggest that our model qualitatively gained a viewpoint similar to that of domain experts. A series of analyses revealed that the data-driven approach using only deep learning has limitations, and the integration of domain knowledge could bring new breakthroughs.

Assessment of sensory loads in civil aviation pilots

Introduction. Performing flights on modern types of aircraft is accompanied by an increase in the role of process automation, changes in the structure of the pilot’s information field, an increase in the number of controlled indicators throughout the flight time and, as a result, an increase in sensory loads that contribute to the development of chronic depletion of functional reserves of the body and fatigue of the pilot, which can cause accidents and plane crashes.The aim of the study is to assess the sensory loads of members of flight crews of civil aviation aircraft when they perform regular operating procedures in a Training center on fully functional complex simulators for training cadets, flight, dispatcher and technical personnel of civil aviation.Materials and methods. A hygienic study was conducted to assess the sensory loads of pilots in the framework of performing simulation flights on full-featured complex simulators in strict step-by-step accordance with the manual for flight operation of the aircraft (33 flights were performed with the participation of 66 pilots aged 30–55 years).Results. The data obtained indicate high levels of sensory loads in pilots, which are 9.5 times higher than the maximum values set by the Management of R 2.2.2006–05 and MI NTP. INT–17.01–2018. For 4 of the 6 evaluated indicators, sensory loads corresponded to class 3.2.Conclusions. The general assessment of the labor intensity class for members of flight crews of civil aviation aircraft on the set of sensory loads corresponds to harmful strenuous work of the 3rd degree (4 indicators of sensory loads with class 3.2). The results of the study are the basis for the approval of new hygienic criteria for establishing a class of working conditions for certain indicators of sensory loads for members of flight crews of the aircraft of the civil aviation (CA) with the addition of their class 3.3. A real assessment of the pilot’s labor intensity should be obtained only on the basis of time-based measurements performed as part of a step-by-step analysis of the aircraft’s flight operation manual.

Analysis of Morphometric Variables of River Espinharas Hydrographic Sub-basin Using Geographic Information System

The relief forms, terrain differences, soil type and flora are the most important indicators for the description of a river basin or drainage network. The present work had as objectives, to perform the morphometric characterization of the hidrographic sub-basin (SBH) of the Espinharas river; analyze the intensity of land use; suggesting mitigating measures in areas with greater intensity of use, with the help of geospatial tools. The sub-basin of the Espinharas river, extends through thirty-one (31) municipalities, covering the states of Paraíba, Rio Grande do Norte and Pernambuco. For the analysis, images were used, from the Digital Elevation Model (DEM), Shuttle Radar Topographic Mission (SRTM), and Landsat 8 satellite images, resolution 30 m, bands 2, 3 and 4. Next, with QGIS software aid 2.18.17 generated the map of slope, mapping of intensity and land use, later calculations of the morphometric variables, and finally elaborated proposals of mitigating measures for the degradation of the sub-basin. The results of the morphometric parameters found for the Sub-basin indicate values of the Compass coefficient of 2.68 and Form factor 0.32, indicating that the sub-basin presents an irregular shape that differs from the figure of a circle, approaching an elongated shape, and thus not conducive to flooding. In relation to the intensity of use, the areas with low intensity class of use have the largest representation in this basin, covering an area of 2,147.98 km² (65.27%). The Espinharas river sub-basin presents several nonconformities of environmental impacts generated mainly by bad planning of use of the area and disrespect to the legislation. In this case, it is necessary for research to support effective public policies that favor less impacting agricultural practices, allowing farmers to provide their livelihoods at the same time, allowing future generations to survive in the semi-arid.

Not Having Been Breastfed May Protect Chinese Women From Developing Deep Infiltrating Endometriosis: Results From Subgroup Analyses of the FEELING Study

Background: This study aimed to investigate potential factors, especially early-life exposures, associated with endometrioma (OMA) and/or deep infiltrating endometriosis (DIE) in Chinese women. Methods: This is a subgroup analyses of the FEELING study, which was a case–control study that investigated the clinical, lifestyle, and environmental factors associated with OMA and/or DIE in China, Russia, and France. In this subgroup analysis, the data for the Chinese participants were further analyzed using logistic regression model. Results: All women (N = 546) had fully completed the questionnaire. The mean age of the participants was 31.8 (range: 18-41) years. Univariable analysis showed that noncyclic chronic pelvic pain, dysmenorrhea intensity class, and whether breastfed during infancy were distributed differently between patients with OMA or DIE and those with no endometriosis (non-EM) or superficial peritoneal endometriosis (SUP; P < .05). Multivariable analysis revealed that not having been breastfed was a protective factor against OMA and DIE (odds ratio [OR] = 0.33, 95% confidence interval [CI]: 0.16-0.69). Further analysis indicated not having been breastfed was a protective factor for DIE compared with non-EM (OR = 0.13, 95% CI: 0.02-0.88) and with OMA + SUP (OR = 0.19, 95% CI: 0.04-0.85) but was not a protective factor for OMA compared with non-EM (OR = 0.66, 95% CI: 0.32-1.36) and with SUP (OR = 0.63, 95% CI: 0.31-1.30). Conclusion: This is the first study suggesting that not having been breastfed might protect against DIE in Chinese women.

PO-135 Application of Wireless Heart rate system in Men's Freestyle

Objective Using heart rate wireless system to continuously track the morning pulse of male freestyle wrestlers after different training loads. The basic value of individual morning pulse and the interval of variation of individual morning pulse after class were established for the elite players. Methods The paper is based on 6 elite male freestyle wrestlers, we continuously test morning pulse after different training load and recovery period，then fill in fatigue questionnaire.We use Spss statistical software to carry out statistical treatment of the experimental data, and analysis the descriptive, difference, correlation and reliability of data. Results Men freestyle athletes base morning pulse at 43-47 / min. After high intensity, the morning pulse increased by about 6%-11%, and the range of change was basically consistent with the change of subjective feeling and heart rate in training class. The average (X) and standard deviation (SD) of the morning pulse in high intensity training class were different from each other. Combined with the results of training diary and fatigue questionnaire, the players appeared body and psychological fatigue and complained that they felt very tired. The X-SD~X SD interval can be used as the early morning pulse range for evaluating athletes' fatigue after high intensity class. The morning pulse and SpO2 was associated with exhaustion of emotional energy, negative evaluation of exercise, decreased sense of achievement, and no significant correlation with heart rate fatigue. However, these three dimensions were significantly related to mental fatigue, r>0.934. Conclusions After high intensity, the morning pulse increased significantly compared with the basic value of morning pulse. Combined with questionnaire survey and training diary feedback, the subjective feeling of body appeared fatigue after high intensity class, which was also consistent with the change of morning pulse. Can be based on individual morning pulse changes to learn about the high intensity class fatigue situation.The morning pulse after high intensity can reach the fatigue interval, which indicates that the training intensity can stimulate the body greatly, and the gradual recovery of the morning pulse can be regarded as the state of whether or not there is overfatigue. If maintain oneself high level all the time, need to adjust training intensity in time. If you can gradually recover close to the basic value, the large-intensity training class can be well adapted to the body.

Índice de anomalias de chuva para diferentes mesorregiões do Estado do Rio Grande do Norte (Rain Anomaly Index for different mesoregions of Rio Grande do Norte state)

Uma das formas de se estudar a precipitação pluvial de determinada região é através de Índices de Anomalia de Chuva (IAC). Nesse contexto, o presente trabalho teve como objetivo analisar o IAC em seis municípios de diferentes mesorregiões do Estado do Rio Grande do Norte, evidenciando os anos de seca e chuva. Mossoró, Apodi, Cruzeta, Santa Cruz, Natal e Ceará Mirim foram os municípios selecionados. Os dados de precipitação pluvial das séries históricas foram obtidos na Empresa de Pesquisa Agropecuária do Rio Grande do Norte (EMPARN), e os cálculos do IAC foram realizados com base na normal climatológica e médias dos 10 anos de máximas e mínimas precipitações. A classificação de anos secos e chuvosos foi feita de acordo com as classes de intensidade do IAC. Dentre os municípios analisados, Natal foi o mais chuvoso, reflexo da precipitação média anual (1656,3 mm), com o IAC variando entre -4,47 e 6,78. Cruzeta apresentou 29 anos com desvios negativos e 25 anos com desvios positivos. Mossoró foi a única cidade a apresentar 2 anos extremamente secos, e teve o IAC variando entre -4,12 e 6,78. Diferentemente de Mossoró, Ceará Mirim teve dois anos extremamente chuvosos, e o IAC variou entre -4,43 e 5,07. Apodi foi a cidades que mais apresentou eventos secos, e a cidade de Santa Cruz foi a única que não apresentou nenhum ano extremamente seco. Concluiu-se que a utilização do IAC garante resultados mais precisos referentes ao grau de variação da chuva em torno da normal climatológica. A B S T R A C TOne way to study the rainfall of the determined region is through Rain Anomaly Indexes (IAC). In this context, the present work had the objective of analyzing the IAC in six municipalities of different mesoregions of the State of Rio Grande do Norte, evidencing the years of drought and rain. Mossoró, Apodi, Cruzeta, Santa Cruz, Natal e Ceará Mirim were the municipalities selected. The rainfall data of the historical series were obtained from the Agricultural Research Company of Rio Grande do Norte (EMPARN), and the calculations of the IAC were performed based on the normal climatologica and averages of 10 years of maximum and minimum precipitations. The classification of dry and wet years was made according to intensity classes of IAC. Among the municipalities analyzed, Natal was the wettest, reflecting the annual average precipitation (1656,3 mm), with the IAC ranging between -4,47 and 6,78. Cruzeta presented 29 years with negative deviations and 25 years with positive deviations. Mossoró was the only city to present two extremely dry years, and had IAC varying between -4,12 e 6,78. Unlike Mossoró, Ceará Mirim had two extremely wet years, and the IAC ranged between -4,43 and 5,07. Apodi was the city that more presented dry events, and the Santa Cruz city was the only one that did not present dry years. It was concluded that the use of IAC guarantees more accurate results regarding the degree of rain variation around the climatological normal.Keywords: rainfall; historical data; intensity class.

Infrared Image Processing for Operating Point Control of Gas Turbine Combustor

The paper presents a technique to estimate the combustor equivalence ratio based on broadband infrared flame images. A static zonal interrogation technique was developed to separate flame images from the combustor background. Histogram characteristics of gray scale images were used to correlate the image properties with flame equivalence ratios. Two histogram properties i.e. average number of pixels at 100 to 200 intensity-class and number of pixels at 150 intensity-classes were used to develop the correlation.

Predicting the Adoption and Maintenance of Exercise Participation Using Self-Efficacy and Previous Exercise Participation Rates

Objectives. To investigate the relationships among self-efficacy, changes in self-efficacy, past exercise participation, future exercise adherence, and exercise program format. Methods. Two-year randomized trial involving subjects (n = 63) participating in an aerobic exercise program. Subjects were randomly assigned to one of three exercise conditions: higher-intensity home-based exercise, higher-intensity class-based exercise, or lower-intensity home-based exercise. Results. Results indicated that baseline self-efficacy and exercise format had significant (p <.02), independent effects on adherence during the adoption and early maintenance phases of exercise behavior. In contrast, in predicting long-term exercise program maintenance, a significant (p <.05) self-efficacy X exercise format interaction indicated that self-efficacy predicted adherence only in the supervised home-based exercise conditions. Results also suggest that baseline self-efficacy, independent of the effect of past adherence, significantly (p <.03) predicted exercise adherence during the adoption phase, but not early maintenance phase, of exercise behavior. Finally, adherence change during the adoption phase of exercise behavior significantly (p <.04) predicted Year-one levels of self-efficacy even after adjusting far the effect of baseline self-efficacy. Conclusions. These results suggest that exercise program format as well as an individual's initial cognitive and behavioral experiences in an exercise program play significant roles in determining exercise adherence.