Collado Rosell, Arturo (2022) Procesamiento estadísticos de señales en aplicaciones de radar meteorológico / Statistical signal proccessing in weather radar application. Tesis Doctoral en Ciencias de la Ingeniería, Universidad Nacional de Cuyo, Instituto Balseiro.
![[img]](/style/images/fileicons/application_pdf.png)
| PDF (Tesis) Español 19Mb |
Resumen en español
El radar meteorológico es un sistema activo de sensado remoto que se utiliza para realizar alertas meteorológicas de corto plazo, contribuyendo a la prevención de pérdidas de indoles humanas y económicas. Su principio de funcionamiento consiste en transmitir energía en forma de ondas electromagnéticas y recibir parte de la energía reflejada por los fenómenos meteorológicos de interés. Dada la naturaleza aleatoria de la señal recibida, se utilizan técnicas de procesamiento estadístico para obtener información útil de los fenómenos meteorológicos bajo estudio. En particular, en los radares Doppler meteorológicos los parámetros de interés lo constituyen los tres momentos de menor orden del espectro de la señal: la potencia, la velocidad Doppler media, y el ancho espectral. Una de las dificultades a tener en cuenta en los radares es que la señal de interés suele encontrarse obscurecida por reflexiones no deseadas, a las cuales se las denomina clutter. Existen diferentes fuentes de clutter, en el caso del radar meteorológico el clutter terrestre, debido a reflexiones producidas sobre el suelo y todo lo que allá se encuentra posee un impacto significativo sobre las estimaciones de los parámetros de interés, por lo que sus efectos se deben eliminar o reducir. Por otro lado, como en todo radar pulsado, el radar meteorológico presenta los problemas de ambigüedades en la determinación del rango y la velocidad Doppler. En el modo de operación convencional, es decir cuando se utiliza un único valor para el intervalo de repetición de pulsos (PRI), aumentar el rango no ambiguo implica disminuir la velocidad Doppler máxima no ambigua y viceversa. La solución más utilizada para lidiar con esta relación de compromiso es alternar el PRI, en general, entre dos valores, a lo que se denomina modo de operación staggered. En esta tesis se abordan los problemas del filtrado de clutter terrestre, la estimación de los momentos espectrales y la clasificanción de la composición de la señal recibida para el radar meteorológico Doppler. Se presentan soluciones tanto para el modo convencional de operación del radar, como para el modo de operación staggered. Inicialmente, se propone una corrección para el ancho espectral del clutter observado con el objetivo de mejorar el filtrado del mismo con aquellos algoritmos que realizan el procesamiento en el dominio del espectro. Posteriormente se introduce el algoritmo Gaussian Model Adaptive Processing No Uniform (GMAP-NU) como solución al filtrado de clutter terrestre y estimación de momentos espectrales de señales adquiridas en el modo de operación staggered. Se muestra que el algoritmo presenta un buen desempeño, comparable al del algoritmo Gaussian Model Adaptive Processing (GMAP) con PRI uniforme. Sin embargo, posee una restricción propia sobre la velocidad Doppler máxima que impide explotar el intervalo Doppler no ambiguo que ofrece el modo staggered y limita su uso desde un punto de vista práctico. Luego, se desarrolla el algoritmo Adaptive Spectral Processing for Staggered Signals (ASPASS), también para el filtrado de clutter terrestre y estimación de los momentos espectrales del fenómeno meteorológico. El mismo se basa en ideas análogas a las de GMAP pero aplicadas a secuencias adquiridas con el modo de operación staggered. Se estudia su desempeño empleando tanto simulaciones numéricas como datos reales adquiridos con el radar argentino RMA-12 situado en el aeropuerto de la ciudad de San Carlos de Bariloche. Los resultados son comparables a los de GMAP-TD, e inclusive mejores, en cuanto a errores relativos y tiempos de cómputo. Por otro lado, se introduce un enfoque diferente para el procesamiento de la señal radar meteorológico que consiste en emplear herramientas de machine learning. En primer lugar, se utilizan redes neuronales artificiales para la estimación de los momentos espectrales tanto en presencia como en ausencia de contribuciones de clutter terrestre en la señal recibida. Se emplea la densidad espectral de potencia (DEP) como entrada a las redes y éstas últimas son entrenadas utilizando datos sintéticos, lo que permite crear bases de datos contemplando una gran diversidad de configuraciones meteorológicas. Este enfoque se aplica a los modos de operación convencional y staggered. En ambos casos, se estudia el desempeño de la respectiva red por medio de simulaciones numéricas y mediciones reales adquiridas con el radar RMA-12. En general, el desempeño es comparable al de algoritmos ampliamente utilizados en la comunidad para resolver este tipo de problemas, lo que pone de manifiesto la versatilidad del método propuesto teniendo en cuenta que el entrenamiento se realiza empleando datos sintéticos. Finalmente, se emplean redes neuronales convolucionales en el problema de clasificación de la composición de la señal recibida, con el objetivo de detectar la presencia de clutter terrestre. Nuevamente, se toma la DEP como entrada a las redes, las mismas son entrenadas utilizando datos sintéticos y el clasificador se aplica para ambos modos de operación del radar. Para cada modo, las redes entrenadas son evaluadas mediante diferentes experimentos por medio de simulaciones numéricas y su funcionamiento es validado a través de mediciones reales adquiridas por los radares RMA-11 y RMA-12. Los resultados muestran que las redes entrenadas poseen tasas de acierto mayor al 90% en la mayor parte de las situaciones estudiadas, y que su desempeño se degrada en configuraciones meteorológicas puntuales, en las cuales es difícil distinguir la contribución de cada componente sobre la DEP resultante.
Resumen en inglés
A weather radar is a remote sensing active device used in short term (nowcasting) meteorological alert systems, allowing the prevention of human and economic losses. Its operating principle consists in transmitting energy through electromagnetic waves and receiving part of the energy reflected by the meteorological objects of interest. Due to the random nature of the received signal, statistical processing techniques are used to obtain useful information of the weather phenomenon under study. In Doppler weather radars, the parameters of interest are the rst three moments of the signal spectrum. One of the difficulties to consider in radars is that the signal of interest is usually obscured by unwanted reflections, known as clutter. There are different clutter sources. In the case of the weather radar, the ground clutter, due to reflections produced by the ground and everything on it, has a huge impact on the estimation of the parameters of interest, therefore its effect must be eliminated or reduced. Also, as in every pulsed radar, the weather radar presents range and Doppler ambiguity problems. In the conventional operating mode, when only one value for the pulse repetition interval (PRI) is used, increasing the maximum unambiguous range implies decreasing the maximum Doppler velocity, and vice versa. The most used solution to deal with this trade-of is to use different PRI values, alternating between two values, denominated satggered operating mode. In this thesis, the ground clutter ltering problem, the spectral moments estimation and the received signal composition classication for the Doppler weather radar are addressed. Solutions are presented for both the conventional and the staggered radar operating modes. The rst result presented is a correction for ground clutter spectral width estimation, which aims at improving the ltering stage in all those algorithms which use the spectral domain to process the signal. Afterwards, a solution to ground clutter ltering and spectral moments estimation for signals acquired in the staggered operating mode is introduced, named GMAP-NU. It is shown that the algorithm presents a performance similar to GMAP for uniform PRI. The main drawback of this algorithm is a restriction in the maximum velocity which prevents it to exploit the entire unambiguous Doppler interval that the staggered mode allows, thus limiting its use for real implementations. Next, a novel ground clutter ltering and spectral moments estimation algorithm is presented, named ASPASS. It is based on the same principle than GMAP but applied to staggered sequences. ASPASS performance is studied both by means of numerical simulations and also using real data acquired with the RMA-12 weather radar, located in San Carlos de Bariloche airport. The results are akin to those obtained with GMAPTD, and outperform this method in terms of relative errors and computing times. Also, a different approach for weather radar signal processing is presented, which relies on machine learning tools. Neural networks are used for weather spectral moments estimation both in the presence and absence of ground clutter. The power spectral density (PSD) is used to feed the neural networks. These are trained using synthetic data, which allows for the creation of huge databases contemplating a great diversity of meteorological situations. This approach is applied to the conventional and the staggered operating modes. In both cases, the neural network performance is studied by means of numerical simulations and real data acquired with the RMA-12 weather radar. In general, the performance is comparable with other widely used algorithms to solve this kind of problems, which reveals the versatility of the proposed method having in mind that the training is performed by means of synthetic data. Finally, convolutional neural networks are used to solve a signal classication problem, with the objective of detecting the presence of ground clutter. Again, the PSD is used to feed neural networks, which are trained using synthetic data and the method is, again, applied to both radar operating modes. For every mode, trained neural networks are tested by means of numerical simulations and their behaviour is validated using real data from RMA-11 and RMA-12 radars. Results show that the trained networks have hit rates greater than 90 % in the majority of the considered situations. The algorithms performance degrades for very particular weather situations, in which it is hard to distinguish the contribution of each component to the resultant PSD.
| Tipo de objeto: | Tesis (Tesis Doctoral en Ciencias de la Ingeniería) |
|---|---|
| Palabras Clave: | Neural network; Redes neuronales; Radar meteorology ; Meteorología con radar; [Ground clutter; Clutter terrestre; Spectral moments; Momentos espectrales; Doppler processing; Procesamiento doppler] |
| Referencias: | [1] Jackson, J. D. Classical electrodynamics, 1999. 1 [2] Skolnik, M. I., et al. Introduction to radar systems. New York, 19621, 2001. 1, 2, 3, 16 [3] Doviak, R. J., Zrnic, D. S. Doppler Radar & Weather Observations. Academic Press, 2014. 1, 2, 4, 14, 18, 19, 20, 21, 22, 25, 26, 27, 29, 55, 67, 96, 108, 112, 143, 151, 156, 177, 183 [4] Marcum, J. A statistical theory of target detection by pulsed radar. IRE Transactions on Information Theory, 6 (2), 59-267, 1960. 2 [5] Bringi, V. N., Chandrasekar, V. Polarimetric Doppler weather radar: principles and applications. Cambridge university press, 2001. 2, 41 [6] Richards, M. A. Fundamentals of radar signal processing. McGraw-Hill Education, 2014. 2, 3, 16, 23 [7] Janssen, L. H., Van Der Spek, G. A. The shape of Doppler spectra from precipitation. IEEE Transactions on Aerospace and electronic systems, (2), 208-219, 1985. 2, 76, 99, 122, 146, 151, 173 [8] Ryzhkov, A. V., Zrnic, D. S. Radar polarimetry for weather observations. Springer, 2019. 3, 17, 18, 19, 20, 95 [9] Rodríguez, A. e. SiNaRaMe: Integración de una red de radares hiro-meteorológicos en Latinoamérica. Rev. Fac. Ciencias Exac, Fis. y Nat., 4 (1), 41-48, 2017. 3 [10] de Elía, R., Vidal, L., Lohigorry, P., Mezher, R., Rugna, M. La red argentina de radares meteorológicos de argentina. Nota Técnica SMN, 39, 2017. 3 [11] Haykin, S. Radar signal processing. IEEE ASSP Magazine, 2 (2), 2-18, 1985. 3 [12] Kelly, E. Performance of an adaptive detection algorithm; rejection of unwanted signals. IEEE Transactions on Aerospace and Electronic Systems, 25 (2), 122-133, 1989. 3 [13] Warde, D. A., Torres, S. Automatic detection and removal of ground clutter contamination on weather radars. En: Proc. 34th Conference on Radar Meteorology. 2009. 3 [14] Hubbert, J., Dixon, M., Ellis, S. Weather radar ground clutter. part ii: Realtime identication and ltering. Journal of Atmospheric and Oceanic Technology, 26 (7), 1181-1197, 2009. 3, 5, 8, 93 [15] Doviak, R. J., Zrnic, D. S., Sirmans, D. S. Doppler weather radar. Proceedings of the IEEE, 67 (11), 1522-1553, 1979. 3, 4 [16] Tahanout, M., Du Chatelet, J. P., et al. An improved M-PRT technique for spectral analysis of weather radar observations. IEEE Transactions on Geoscience and Remote Sensing, 53 (10), 5572-5582, 2015. 4 [17] Zrinc, D., Mahapatra, P. Two methods of ambiguity resolution in pulse Doppler weather radars. IEEE Transactions on Aerospace and Electronic Systems, (4), 470-483, 1985. 4, 7, 25, 38, 39, 67, 177 [18] Sirmans, D., Bumgarner, B. Numerical comparison of ve mean frequency estimators. Journal of Applied Meteorology (1962-1982), 14 (6), 991-1003, 1975. 4 [19] Mahapatra, P. R., Zrnic, D. S. Practical algorithms for mean velocity estimation in pulse Doppler weather radars using a small number of samples. IEEE transactions on geoscience and remote sensing, (4), 491-501, 1983. 4, 5 [20] Zrnic, D. Spectral moment estimates from correlated pulse pairs. IEEE Transactions on Aerospace and Electronic Systems, (4), 344-354, 1977. 5 [21] Chornoboy, E. S. Optimal mean velocity estimation for Doppler weather radars. IEEE transactions on geoscience and remote sensing, 31 (3), 575-586, 1993. 5 [22] Pinsky, M., i Ventura, J. F., Otto, T., Sterkin, A., Khain, A., Russchenberg, H. W. Application of a simple adaptive estimator for an atmospheric Doppler radar. IEEE transactions on geoscience and remote sensing, 49 (1), 115-127, 2010. 5 [23] Groginsky, H. L., Glover, K. Weather radar canceller design. En: 19th Conference on Radar Meteorology, pags. 192-198. 1980. 5 [24] Siggia, A., Passarelli, R. Gaussian model adaptive processing (GMAP) for improved ground clutter cancellation and moment calculation. En: Proc. ERAD, tomo 2, pags. 421-424. 2004. 5, 29, 68, 73 [25] Nguyen, C. M., Chandrasekar, V. Gaussian model adaptive processing in time domain (GMAP-TD) for weather radars. Journal of Atmospheric and Oceanic Technology, 30 (11), 2571-2584, 2013. 6, 7, 34, 36, 41, 42, 68 [26] Torres, S. M.,Warde, D. A. Ground clutter mitigation for weather radars using the autocorrelation spectral density. Journal of Atmospheric and Oceanic Technology, 31 (10), 2049-2066, 2014. 5, 6 [27] Warde, D. A., Torres, S. M. The autocorrelation spectral density for Dopplerweather- radar signal analysis. IEEE transactions on geoscience and remote sensing, 52 (1), 508-518, 2013. 6 [28] Nguyen, C. M., Moisseev, D. N., Chandrasekar, V. A parametric time domain method for spectral moment estimation and clutter mitigation for weather radars. Journal of Atmospheric and Oceanic Technology, 25 (1), 83-92, 2008. 6 [29] Sachidananda, M., Zrnic, D. An improved clutter ltering and spectral moment estimation algorithm for staggered PRT sequences. Journal of Atmospheric and Oceanic Technology, 19 (12), 2009-2019, 2002. 7, 8, 40, 75 [30] Pascual, J. P., Cogo, J., Collado Rosell, A., Areta, J. Multipulse Processing Algorithm for Improving Mean Velocity Estimation in Weather Radar. IEEE Transactions on Geoscience and Remote Sensing, 2021. 7, 11, 135, 183 [31] Moisseev, D. N., Nguyen, C. M., Chandrasekar, V. Clutter suppression for staggered PRT waveforms. J. Atmos. Oceanic Technol., 25 (12), 2209-2218, 2008. 7 [32] Torres, S. M., Warde, D. A. Staggered-PRT sequences for Doppler weather radars. part i: Spectral analysis using the autocorrelation spectral density. J. Atmos. Oceanic Technol., 34 (1), 51-63, 2017. 7 [33] Warde, D. A., Torres, S. M. Staggered-PRT sequences for Doppler weather radars. part ii: Ground clutter mitigation on the NEXRAD network using the CLEAN-AP lter. J. Atmos. Oceanic Technol., 34 (3), 703-716, 2017. 7 [34] Sachidananda, M., Zrnic, D. Clutter ltering and spectral moment estimation for Doppler weather radars using staggered pulse repetition time (PRT). Journal of Atmospheric and Oceanic Technology, 17 (3), 323-331, 2000. 7, 67, 70, 183 [35] Li, Y., Zhang, G., Doviak, R. J. Ground clutter detection using the statistical properties of signals received with a polarimetric radar. IEEE Transactions on Signal Processing, 62 (3), 597-606, 2013. 8 [36] L-Garcia, A. Probability, Statistics and Random Processes for Electrical Engineering, 2008. 8, 56 [37] Golbon-Haghighi, M.-H., Zhang, G., Li, Y., Doviak, R. J. Detection of ground clutter from weather radar using a dual-polarization and dual-scan method. Atmosphere, 7 (6), 83, 2016. 8 [38] Golbon-Haghighi, M.-H., Zhang, G., Doviak, R. J. Ground clutter detection for weather radar using phase fructuation index. IEEE Transactions on Geoscience and Remote Sensing, 57 (5), 2889-2895, 2018. 8 [39] Collado Rosell, A. C., Pascual, J. P., Areta, J. Spectrum width correction for clutter mitigation in weather radar. En: 2019 XVIII Workshop on Information Processing and Control (RPIC), pags. 135-139. IEEE, 2019. 9, 53 [40] Collado Rosell, A., Areta, J., Pascual, J. Ground clutter ltering for weather radar using staggered pulse repetition time. En: 2018 IEEE Biennial Congress of Argentina (ARGENCON), pags. 1-6. IEEE, 2018. 9, 54 [41] Collado Rosell, A., Pascual, J. P., Areta, J. Adaptive spectral processing algorithm for staggered signals in weather radars. IET Radar, Sonar & Navigation, 14 (11), 1659-1670, 2020. 9, 68 [42] Collado Rosell, A., Cogo, J., Areta, J. A., Pascual, J. P. Doppler processing in weather radar using deep learning. IET Signal Processing, 14 (9), 672-682, 2020. 10, 96 [43] Meikle, H. Modern radar systems. Artech House, 2008. 13 [44] Ali, B. S. Aircraft Surveillance Systems: Radar Limitations and the Advent of the Automatic Dependent Surveillance Broadcast. Routledge, 2017. 13 [45] Moreira, A., Prats-Iraola, P., Younis, M., Krieger, G., Hajnsek, I., Papathanassiou, K. P. A tutorial on synthetic aperture radar. IEEE Geoscience and remote sensing magazine, 1 (1), 6-43, 2013. 13 [46] Papoulis, A., Unnikrishna, S. P. Probability, random variables and stochastic processes. McGraw-Hill, 2002. 19, 26, 55, 56, 185 [47] Dutta, A., Chandrasekar, V. Detection, Analysis and Mitigation of Sea Clutter in Polarimetric Weather Radar. En: 2019 United States National Committee of URSI National Radio Science Meeting (USNC-URSI NRSM), pags. 1-2. 2019. 22 [48] Hailong, W., Shouyuan, D., Xu, W., Zhao, S. Sea Clutter Recognition Based on Dual-Polarization Weather Radar. En: 2019 International Conference on Meteorology Observations (ICMO). 2019. 22 [49] Uysal, F., Selesnick, I., Isom, B. M. Mitigation of Wind Turbine Clutter for Weather Radar by Signal Separation. IEEE Transactions on Geoscience and Remote Sensing, 54 (5), 2925-2934, 2016. 22, 183 [50] Hood, K., Torres, S., Palmer, R. Automatic Detection of Wind Turbine Clutter for Weather Radars. Journal of Atmospheric and Oceanic Technology, 27 (11), 1868-1880, 11 2010. 22, 183 [51] Radhakrishna, B., Fabry, F., Kilambi, A. Fuzzy Logic Algorithms to Identify Birds, Precipitation, and Ground Clutter in S-Band Radar Data Using Polarimetric and Nonpolarimetric Variables. Journal of Atmospheric and Oceanic Technology, 36 (12), 2401-2414, 12 2019. 22, 183 [52] Porat, B. A course in digital signal processing. Wiley, 1997. 24, 54, 70, 73, 84, 86 [53] Welch, P. The use of fast fourier transform for the estimation of power spectra: a method based on time averaging over short, modied periodograms. IEEE Transactions on audio and electroacoustics, 15 (2), 70-73, 1967. 30, 70 [54] Hildebrand, P. H., Sekhon, R. Objective determination of the noise level in Doppler spectra. Journal of Applied Meteorology, 13 (7), 808-811, 1974. 30, 75, 80 [55] Cho, J. Y., Chornoboy, E. S. Multi-PRI signal processing for the terminal Doppler weather radar. part i: Clutter ltering. Journal of Atmospheric and Oceanic Technology, 22 (5), 575-582, 2005. 42 [56] Zrnic, D., et al. Simulation of weatherlike Doppler spectra and signals. Journal of Applied Meteorology, 14 (4), 619-620, 1975. 43 [57] Mitchell, T. M. Does machine learning really work? AI magazine, 18 (3), 11-11, 1997. 44 [58] Goodfellow, I., Bengio, Y., Courville, A., Bengio, Y. Deep learning, tomo 1. MIT press Cambridge, 2016. 44, 45, 46, 48, 49, 50, 51, 95 [59] Han, J., Moraga, C. The influence of the sigmoid function parameters on the speed of backpropagation learning. En: International Workshop on Artificial Neural Networks, pags. 195-201. Springer, 1995. 47 [60] Kalman, B. L., Kwasny, S. C. Why tanh: choosing a sigmoidal function. En: [Proceedings 1992] IJCNN International Joint Conference on Neural Networks, tomo 4, pags. 578{581. IEEE, 1992. 47 [61] Nair, V., Hinton, G. E. Rectied linear units improve restricted boltzmann machines. En: Proce of the 27th interna confe on machi learn (ICML-10), pags. 807-814. 2010. 47 [62] Dunne, R. A., Campbell, N. A. On the pairing of the softmax activation and crossentropy penalty functions and the derivation of the softmax activation function. En: Proc. 8th Aust. Conf. on the Neural Networks, Melbourne, tomo 181, pag. 185. Citeseer, 1997. 47 [63] Zou, F., Shen, L., Jie, Z., Zhang, W., Liu, W. A sucient condition for convergences of adam and rmsprop. En: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pags. 11127-11135. 2019. 48 [64] Kingma, D. P., Ba, J. Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980, 2014. 48, 102 [65] Rumelhart, D. E., Durbin, R., Golden, R., Chauvin, Y. Backpropagation: The basic theory. Backpropagation: Theory, architectures and applications, pags. 1-34, 1995. 49 [66] Apostol, T. M. Calculus, Volume 1. John Wiley & Sons, 1991. 49 [67] Srivastava, N., Hinton, G., Krizhevsky, A., Sutskever, I., Salakhutdinov, R. Dropout: a simple way to prevent neural networks from overtting. The journal of machine learning research, 15 (1), 1929-1958, 2014. 51 [68] Ioe, S., Szegedy, C. Batch normalization: Accelerating deep network training by reducing internal covariate shift. En: International conference on machine learning, pags. 448-456. PMLR, 2015. 51 [69] Dolev, D., Reischuk, R., Strong, H. R. Early stopping in Byzantine agreement. Journal of the ACM (JACM), 37 (4), 720-741, 1990. 51 [70] Russakovsky, O., Deng, J., Su, H., Krause, J., Satheesh, S., Ma, S., et al. ImageNet Large Scale Visual Recognition Challenge. International Journal of Computer Vision (IJCV), 115 (3), 211-252, 2015. 51 [71] Ziemer, R. E., Tranter, W. H. Principles of communications. John Wiley & Sons, 2014. 56, 57 [72] Bjorck, A. Numerical methods for least squares problems. SIAM, 1996. 58 [73] Jenq, Y.-C. Perfect reconstruction of digital spectrum from nonuniformly sampled signals. IEEE Transactions on Instrumentation and Measurement, 46 (3), 649-652, 1997. 61 [74] Banjanin, Z. B., Zrnic, D. S. Clutter rejection for Doppler weather radars which use staggered pulses. IEEE Transactions on Geoscience and Remote Sensing, 29 (4), 610-620, 1991. 68 [75] Gregory Meymaris, N. Alternative approaches to staggered PRT clutter ltering. En: 26th Conference on Interactive Information and Processing Systems. 2010. 68 [76] Saltiko, E., Cho, J. Y., Tristant, P., Huuskonen, A., Allmon, L., Cook, R., et al. The threat to weather radars by wireless technology. Bulletin of the American Meteorological Society, 97 (7), 1159-1167, 2016. 88 [77] Ice, R. L., Rhoton, R., Krause, J., Saxion, D., Boydstun, O., Heck, A., et al. Automatic clutter mitigation in the WSR-88D, design, evaluation, and implementation. En: 34th Conference on Radar Meteorology, Williamsburg, VA. 2009. 93 [78] Ivic, I. R., Curtis, C., Torres, S. M. Radial-based noise power estimation for weather radars. Journal of Atmospheric and Oceanic Technology, 30 (12), 2737- 2753, 2013. 94 [79] Kononenko, I. Machine learning for medical diagnosis: history, state of the art and perspective. Artificial Intelligence in medicine, 23 (1), 89-109, 2001. 95 [80] Zhang, L., Tan, J., Han, D., Zhu, H. From machine learning to deep learning: progress in machine intelligence for rational drug discovery. Drug discovery today, 22 (11), 1680-1685, 2017. 95 [81] Gyor, L., Ottucsak, G., Walk, H. Machine learning for nancial engineering, tomo 8. World Scientic, 2012. 95 [82] Daily, M., Medasani, S., Behringer, R., Trivedi, M. Self-driving cars. Computer, 50 (12), 18-23, 2017. 95 [83] Islam, T., Rico-Ramirez, M. A., Han, D., Srivastava, P. K. Articial intelligence techniques for clutter identification with polarimetric radar signatures. Atmospheric Research, 109, 95-113, 2012. 95 [84] Wang, H., Ran, Y., Deng, Y., Wang, X. Study on deep-learning-based identification of hydrometeors observed by dual polarization Doppler weather radars. EURASIP Journal on Wireless Communications and Networking, 2017 (1), 1-9, 2017. 95 [85] Li, H., Ren, J., Han, J., Fan, Y. Ground clutter suppression method based on FNN for dual-polarisation weather radar. The Journal of Engineering, 2019 (19), 6043-6047, 2019. 95 [86] Zadeh, L. A. Fuzzy logic. Computer, 21 (4), 83-93, 1988. 95 [87] Abadi, M., Agarwal, A., Barham, P., Brevdo, E., Chen, Z., Citro, C., et al. TensorFlow: Large-Scale Machine Learning on Heterogeneous Systems, 2015. URL https://www.tensorflow.org/, software available from tensor flow.org. 101, 153 [88] Chollet, F. Keras, 2015. URL https://github.com/fchollet/keras. 101, 153 [89] Pan, S. J., Yang, Q. A survey on transfer learning. IEEE Transactions on know- ledge and data engineering, 22 (10), 1345-1359, 2009. 123, 183 [90] Walke, B. H., Mangold, S., Berlemann, L. IEEE 802 wireless systems: protocols, multi-hop mesh/relaying, performance and spectrum coexistence. John Wiley & Sons, 2007. 182 [91] Galati, G., Pavan, G. Computer simulation of weather radar signals. Simulation Practice and Theory, 3 (1), 17-44, 1995. 184 |
| Materias: | Ingeniería en telecomunicaciones > Procesamiento de señales |
| Divisiones: | Gcia. de área de Investigación y aplicaciones no nucleares > Laboratorio de investigación aplicada en Telecomunicaciones |
| Código ID: | 1068 |
| Depositado Por: | Tamara Cárcamo |
| Depositado En: | 12 Jul 2022 15:13 |
| Última Modificación: | 12 Jul 2022 15:13 |
Personal del repositorio solamente: página de control del documento
