Accelerated Mass Spectrometry Cyclotron System Using AI

摘要

The present invention is a cyclotron-based accelerator mass spectrometry system using artificial intelligence in order to shorten the mass analysis time by increasing the mass decomposition efficiency of a cyclotron-based accelerator mass spectrometry system in which mass is decomposed by accelerating in AC form. The cyclotron-based accelerator mass spectrometry system using the artificial intelligence is an ion source that generates charged particles, a first ion beam delivery unit connected to the ion source and receiving charged particles generated from the ion source and emitting them into the cyclotron, the An RF amplifier that supplies RF to the inside of the cyclotron to accelerate charged particles, is connected to the cyclotron, and is connected to the second ion beam delivery unit and the second ion beam delivery unit to input the charged particles accelerated inside the cyclotron. 2 A two-pole electromagnet that receives the accelerated charged particles from an ion beam delivery unit and draws out charged particles to be analyzed, a charged particle detection unit that is connected to the two-pole electromagnet and detects a ratio of charged particles to be analyzed, the It is connected to the RF amplifier and the charged particle detection unit, generates an RF amplifier driving signal that drives the RF amplifier and delivers it to the RF amplifier, and generates a charged particle detection unit driving signal that drives the charged particle detection unit, and the charged particle The control unit transmitted to the detection unit, connected to the RF amplifier and the charged particle detection unit, receive and accumulate the RF frequency from the RF amplifier and the ratio of charged particles to be analyzed from the charged particle detection unit, and based on the accumulated information, an artificial neural network And an artificial intelligence unit that predicts the average value of the ratio of the charged particles to be analyzed according to the RF frequency by using the based artificial intelligence learning method, and generates and stores the pattern of the charged particles to be analyzed according to the RF frequency.