Due to limited size and imperfect of the optical components in a spectrometer, aberration has inevitably been brought into two-dimensional multi-fiber spectrum image in LAMOST, which leads to obvious spacial variation of the point spread functions (PSFs). Consequently, if spatial variant PSFs are estimated directly , the huge storage and intensive computation requirements result in deconvolutional spectral extraction method become intractable. In this paper, we proposed a novel method to solve the problem of spatial variation PSF through image aberration correction. When CCD image aberration is corrected, PSF, the convolution kernel, can be approximated by one spatial invariant PSF only. Specifically, machine learning techniques are adopted to calibrate distorted spectral image, including Total Least Squares (TLS) algorithm, intelligent sampling method, multi-layer feed-forward neural networks. The calibration experiments on the LAMOST CCD images show that the calibration effect of proposed method is effectible. At the same time, the spectrum extraction results before and after calibration are compared, results show the characteristics of the extracted one-dimensional waveform are more close to an ideal optics system, and the PSF of the corrected object spectrum image estimated by the blind deconvolution method is nearly central symmetry, which indicates that our proposed method can significantly reduce the complexity of spectrum extraction and improve extraction accuracy.

Discovering pulsars is a significant and meaningful research topic in the field of radio astronomy. With the advent of astronomical instruments such as he Five-hundred-meter Aperture Spherical Telescope (FAST) in China, data volumes and data rates are exponentially growing. This fact necessitates a focus on artificial intelligence (AI) technologies that can perform the automatic pulsar candidate identification to mine large astronomical data sets. Automatic pulsar candidate identification can be considered as a task of determining potential candidates for further investigation and eliminating noises of radio frequency interferences or other non-pulsar signals. It is very hard to raise the performance of DCNN-based pulsar identification because the limited training samples restrict network structure to be designed deep enough for learning good features as well as the crucial class imbalance problem due to very limited number of real pulsar samples. To address these problems, we proposed a framework which combines deep convolution generative adversarial network (DCGAN) with support vector machine (SVM) to deal with imbalance class problem and to improve pulsar identification accuracy. DCGAN is used as sample generation and feature learning model, and SVM is adopted as the classifier for predicting candidate's labels in the inference stage. The proposed framework is a novel technique which not only can solve imbalance class problem but also can learn discriminative feature representations of pulsar candidates instead of computing hand-crafted features in preprocessing steps too, which makes it more accurate for automatic pulsar candidate selection. Experiments on two pulsar datasets verify the effectiveness and efficiency of our proposed method.

The Commensal Radio Astronomy Five-hundred-meter Aperture Spherical radio Telescope (FAST) Survey (CRAFTS) utilizes the novel drift-scan commensal survey mode of FAST and can generate billions of pulsar candidate signals. The human experts are not likely to thoroughly examine these signals, and various machine sorting methods are used to aid the classification of the FAST candidates. In this study, we propose a new ensemble classification system for pulsar candidates. This system denotes the further development of the pulsar image-based classification system (PICS), which was used in the Arecibo Telescope pulsar survey, and has been retrained and customized for the FAST drift-scan survey. In this study, we designed a residual network model comprising 15 layers to replace the convolutional neural networks (CNNs) in PICS. The results of this study demonstrate that the new model can sort >96% of real pulsars to belong the top 1% of all candidates and classify >1.6 million candidates per day using a dual--GPU and 24--core computer. This increased speed and efficiency can help to facilitate real-time or quasi-real-time processing of the pulsar-search data stream obtained from CRAFTS. In addition, we have published the labeled FAST data used in this study online, which can aid in the development of new deep learning techniques for performing pulsar searches.

We report the detections of a giant planet (MARVELS-7b) and a brown dwarf candidate (MARVELS-7c) around the primary star in the close binary system, HD 87646. It is the first close binary system with more than one substellar circum-primary companion discovered to the best of our knowledge. The detection of this giant planet was accomplished using the first multi-object Doppler instrument (KeckET) at the Sloan Digital Sky Survey (SDSS) telescope. Subsequent radial velocity observations using ET at Kitt Peak National Observatory, HRS at HET, the "Classic" spectrograph at the Automatic Spectroscopic Telescope at Fairborn Observatory, and MARVELS from SDSS-III confirmed this giant planet discovery and revealed the existence of a long-period brown dwarf in this binary. HD 87646 is a close binary with a separation of $\sim22$ AU between the two stars, estimated using the Hipparcos catalogue and our newly acquired AO image from PALAO on the 200-inch Hale Telescope at Palomar. The primary star in the binary, HD 87646A, has Teff = 5770$\pm$80K, log(g)=4.1$\pm$0.1 and [Fe/H] = $-0.17\pm0.08$. The derived minimum masses of the two substellar companions of HD 87646A are 12.4$\pm$0.7M$_{\rm Jup}$ and 57.0$\pm3.7$M$_{\rm Jup}$. The periods are 13.481$\pm$0.001 days and 674$\pm$4 days and the measured eccentricities are 0.05$\pm$0.02 and 0.50$\pm$0.02 respectively. Our dynamical simulations show the system is stable if the binary orbit has a large semi-major axis and a low eccentricity, which can be verified with future astrometry observations.

Exoplanet transit and Doppler surveys discover many binary stars during their operation that can be used to conduct a variety of ancillary science. Specifically, eclipsing binary stars can be used to study the stellar mass-radius relationship and to test predictions of theoretical stellar evolution models. By cross-referencing 24 binary stars found in the MARVELS Pilot Project with SuperWASP photometry, we find two new eclipsing binaries, TYC 0272-00458-1 and TYC 1422-01328-1, which we use as case studies to develop a general approach to eclipsing binaries in survey data. TYC 0272-00458-1 is a single-lined spectroscopic binary for which we calculate a mass of the secondary and radii for both components using reasonable constraints on the primary mass through several different techniques. For a primary mass of M_1 = 0.92 +/- 0.1 M_solar, we find M_2 = 0.610 +/- 0.036 M_solar, R_1 = 0.932 +/- 0.076 R_solar and R_2 = 0.559 +/- 0.102 R_solar, and find that both stars have masses and radii consistent with model predictions. TYC 1422-01328-1 is a triple-component system for which we can directly measure the masses and radii of the eclipsing pair. We find that the eclipsing pair consists of an evolved primary star (M_1 = 1.163 +/- 0.034 M_solar, R_1 = 2.063 +/- 0.058 R_solar) and a G-type dwarf secondary (M_2 = 0.905 +/- 0.067 M_solar, R_2 = 0.887 +/- 0.037 R_solar). We provide the framework necessary to apply this analysis to much larger datasets.

We present a new short-period brown dwarf candidate around the star TYC 1240-00945-1. This candidate was discovered in the first year of the Multi-object APO Radial Velocity Exoplanets Large-area Survey (MARVELS), which is part of the third phase of the Sloan Digital Sky Survey (SDSS-III), and we designate the brown dwarf as MARVELS-1b. MARVELS uses the technique of dispersed fixed-delay interferometery to simultaneously obtain radial velocity measurements for 60 objects per field using a single, custom-built instrument that is fiber fed from the SDSS 2.5-m telescope. From our 20 radial velocity measurements spread over a ~370 d time baseline, we derive a Keplerian orbital fit with semi-amplitude K=2.533+/-0.025 km/s, period P=5.8953+/-0.0004 d, and eccentricity consistent with circular. Independent follow-up radial velocity data confirm the orbit. Adopting a mass of 1.37+/-0.11 M_Sun for the slightly evolved F9 host star, we infer that the companion has a minimum mass of 28.0+/-1.5 M_Jup, a semimajor axis 0.071+/-0.002 AU assuming an edge-on orbit, and is probably tidally synchronized. We find no evidence for coherent instrinsic variability of the host star at the period of the companion at levels greater than a few millimagnitudes. The companion has an a priori transit probability of ~14%. Although we find no evidence for transits, we cannot definitively rule them out for companion radii ~<1 R_Jup.

We report the discovery of a low-mass companion orbiting the metal-rich, main sequence F star TYC 2949-00557-1 during the MARVELS (Multi-object APO Radial Velocity Exoplanet Large-area Survey) Pilot Project. The host star has an effective temperature T_eff = 6135 +/- 40 K, log(g) = 4.4 +/- 0.1 and [Fe/H] = 0.32 +/- 0.01, indicating a mass of M = 1.25 +/- 0.09 M_\odot and R = 1.15 +/- 0.15 R_\odot. The companion has an orbital period of 5.69449 +/- 0.00023 days and straddles the hydrogen burning limit with a minimum mass of 64 M_J, and may thus be an example of the rare class of brown dwarfs orbiting at distances comparable to those of "Hot Jupiters." We present relative photometry that demonstrates the host star is photometrically stable at the few millimagnitude level on time scales of hours to years, and rules out transits for a companion of radius greater than 0.8 R_J at the 95% confidence level. Tidal analysis of the system suggests that the star and companion are likely in a double synchronous state where both rotational and orbital synchronization have been achieved. This is the first low-mass companion detected with a multi-object, dispersed, fixed-delay interferometer.

We report the detection of the first extrasolar planet, ET-1 (HD 102195b), using the Exoplanet Tracker (ET), a new generation Doppler instrument. The planet orbits HD 102195, a young star with solar metallicity that may be part of the local association. The planet imparts radial velocity variability to the star with a semiamplitude of $63.4\pm2.0$ m s$^{-1}$ and a period of 4.11 days. The planetary minimum mass ($m \sin i$) is $0.488\pm0.015$ $M_J$.