Artificial intelligence sheath "sword" refers to exoplanets - The mini-solar system with 8 planets "shows" The solar system finally has an evenly matched "opponent." NASA announced at 2 am Beijing time on December 15th that Google’s artificial intelligence analyzed the data obtained by Kepler and discovered the 8th orbit around the star Kepler-90. The planet Kepler-90i enables the Kepler-90 to keep pace with the Sun and together become the star system that currently possesses the most known planets (8). Artificial intelligence has contributed to this, although previous machine learning technology has been used to search the Kepler telescope database, but the latest research shows that neural networks technology has great potential in capturing the weakest signals emitted by distant objects. Chess rivals the mini solar system "surfaced" Kepler-90 is located 2545 light years away from Earth and is located in Draco. The newly discovered Kepler 90i is a rocky planet that is 1.3 times the size of Earth. Because it is closer to its main star, the orbital period is only 14.4 days, and the average temperature is believed to exceed 800 degrees Fahrenheit (about 426.7 degrees Celsius). One of the heads of the research, Andrew Vanderberger of the University of Texas at Austin, said: "The Kepler-90 system is like a mini version of the solar system: the smaller planets are on the inside; the larger planets are on the outside, but only between them. "Intimate." Nevertheless, it is no better than other planetary systems in bringing up life." Kepler-90i is not the only exoplanet that the latest study "screened out". They also found the Kepler-80 system's sixth planet, Kepler-80g, which is similar in size to the Earth, and its neighboring 4 The planets form an ultra-stable system similar to the TRAPPIST-1 system with seven planets previously discovered. Paul Hertz, head of NASA's astrophysics department, said: "As we expected, in Kepler's telescope data that has been analyzed previously, there are still exciting new discoveries waiting for suitable new tools or new technologies. Reveal it. This discovery shows that these data will be treasures for innovative research for many years to come." Daxian Shenwei neural network divulges "the fish that escapes the net" Christoph Sand, senior software engineer at Google Artificial Intelligence (GoogleAI), first proposed using neural network technology to analyze Kepler's data. He believes that with the continuous upgrading of data acquisition technology, astronomy is slowly falling into the predicament of being swamped with massive data like other disciplines. Therefore, he has a keen interest in using computer technology to search for exoplanets. He said: "When there is too much data, artificial analysis is difficult to parry, it is the time of machine learning 'big show'." For up to four years of operation, the data collected by the Kepler Telescope contained 35,000 suspected planetary signals. Although scientists used artificial methods to screen out the most likely signals, the weakest signals tend to become The fish." The salad and Van Der Berger conjecture that more interesting exoplanet discoveries may be lurking in these data. Thus, the training of artificial intelligence by the salad and Vanderberger found exoplanets based on the brightness data obtained from the Kepler telescope. Inspired by the connection between neurons in the human brain, their artificial neural network technology screened Kepler's data and discovered the previously overlooked 8th planet around Kepler-90 spinning. signal. Exoplanets passing through the front of the star cause the star's brightness to drop slightly. Kepler's telescope records this down signal, which allows scientists to infer the existence of the planet and the physical properties of the planet. This is called The method of sunday. Van der Burg said: "Although we have found a lot of false signals, but it may also hide the real planet. This is like looking through the small stones for jewelry, the better the sieve, the possibility of finding the gem is certainly greater." The researchers plan to next allow the neural network to analyze the brightness data of more than 150,000 stars that Kepler has searched. "Heilongjiang" Kepler wrote a beautiful chapter In addition to artificial intelligence, the most important discovery of this "big hero" is, of course, the Keplerian Telescope. The Kepler telescope was launched in 2009. The initial mission was to scan a small area of ​​the sky. The mission of the mission continued until 2016. However, people are not as good as the heavens. On May 11, 2013, the telescope experienced a critical failure, which made it impossible to accurately locate the original sky area. NASA's geniuses will be wrong, adjust the angle of the solar windsurfing, and use solar radiation light pressure to help the Kepler telescope achieve balance. But everything has a price: The telescope must keep a certain angle with the sun, so the angle that originally scanned a small area of ​​the cygnus was transferred to the ecliptic plane. However, Kepler's telescope is destined to write a legendary, it is "dead and alive," leaving its own mark in the vast history of the universe. From 2014, the Kepler telescope entered K2, the Kepler mission 2. On June 19 this year, NASA released a new set of data that included observations from the Kepler Telescope. It has now discovered 4034 candidate exoplanets, 2335 of which have been confirmed, and 49 are terrestrial planets. The stars have been confirmed and are all on the livable side. Perhaps the most attractive is the world with the serial number "KOI-7711". The planet is 1.3 times the size of the Earth. It revolves around a sun-like star in Earth-like orbit, and is almost the twin brother of Earth. Of course, more details about the world need to wait for a new generation of space telescopes to be unveiled.