- At the Ligo Observatory, deep loop shaping reduces noise 30-100 times.
- This method helps astronomers detect hundreds of collisions between black holes and neutron stars in an annual period.
- This technique was successfully tested on a real-life Ligo system in Louisiana.
Deep loop shaping
Researchers at Google Deepmind have developed a new AI method to improve control of gravitational wave observatory. This method is called the formation of deep loops and helps astronomers better understand the dynamics and formation of the universe.
Deep loop shaping reduces noise and improves control in the station's feedback system. This helps stabilize the components used to measure gravitational waves. Gravitational waves are small ripples in the structure of space and time created by events such as the merger of neutron stars and black holes.
This method was developed in collaboration with Caltech and Ligo (gravitational wave observatory for laser interferometers), run by GSSI (Gran Sasso Science Institute). Researchers have proven that this method works at an observatory in Livingston, Louisiana.
Ligo measures extremely accurately
Ligo measures the properties and origins of gravitational waves with incredible accuracy. However, slight vibrations can disrupt measurements, even from waves crashing 100 miles away on the Gulf of Mexico coast. To function, Ligo relies on thousands of control systems that keep all parts in near perfect alignment and adapt to environmental disorders with continuous feedback.
Deep loop shaping reduces the noise level of the most unstable and difficult feedback loops in Ligo 30-100 times. This improves the stability of the interferometer's sensitive mirrors. Applying this method to all of Ligo's mirror control loops will help astronomers detect and collect about hundreds of events a year.
Observatory uses laser light interference to measure the properties of gravitational waves. By studying these properties, scientists can understand what they cause and where they came from. The observatory's laser reflects a mirror located four kilometers away, housed in the world's largest vacuum chamber.
Disturbance from minimal vibration
As gravitational waves pass through Ligo's two 4-kilometer arms it distorts the space between them, changing the distance between the mirrors at both ends. These small differences in length are measured using optical interference for accuracy of 10^-19 meters. This is one tenth of the size of a proton. For such small measurements, the Ligo detector mirror must be kept very quiet, isolated from environmental disturbances.
This requires one system for passive mechanical separation and another control system to actively suppress vibrations. Less control will swing the mirror and make it impossible to measure anything. However, if there is too much control, instead of actually suppressing the vibrations of the system, it amplifies the signal and stirs up a specific frequency range.
These vibrations, known as “control noise,” are serious obstacles to improving Ligo's ability to look into space. The researchers have designed deep loop shaping to remove controllers as a meaningful source of noise, beyond traditional methods.
Reinforcement learning with frequency domain rewards
Deep Loop Shaping uses reinforcement learning methods with frequency domain rewards, outperforming cutting-edge feedback control performance. In the simulated Ligo environment, researchers trained controllers to avoid amplification of noise in the observation zones used to measure gravitational waves.
Through repeated interactions guided by frequency domain rewards, the controller learns to suppress control noise in the observation band. The controller learns to stabilize the mirror without adding any harmful control noise, lowering the noise level by more than 10 times.
Successful tests on real hardware
Researchers tested the controller on a Real Rigo system in Livingston, Louisiana. They found it works well in hardware as well as in simulations. The results show that deep loop shaping controls noise up to 30-100 times better than existing controllers.
This method is the first meaningful source of noise in LIGO to eliminate the most unstable and difficult feedback loops. In repeated experiments, researchers confirmed that the controllers stabilized the station's system for a long period of time.
Applying deep loop shaping across Ligo's mirror control system can eliminate noise from the control system itself. This opens up a way to expand the cosmological scope of the observation deck. In addition to improving how existing gravitational wave observatory is measured, researchers, who are farther and sources of dimming, hope to have an impact on future observatory designs, both on Earth and in space.
Wall-Y
Wall-Y is an AI bot created with Claude. learn more About Wall-Y and how we develop her. You can find her news here.
You can chat wall gpt About this news article and factual optimism
