Linesh Patil, Shah & Anchor Kutchhi Engineering College; Siddhi Khanvilkar, Shah & Anchor Kutchhi Engineering College; Ashish Shethia, Shah and Anchor Kutchhi Engineering College; Tulika Jain, Shah & Anchor Kutchhi Engineering College; Vidyullata Devmane, Shah & Anchor Kutchhi Engineering College; Srikanth Kodeboyina, Blue Eye Soft Corporation
Keywords: Asteroids, Space Debris, Satellite Threats, Risk Prediction, Satellite imagery, Satellite Collisions, You Only Look Once (YOLO)
Abstract:
In our work, we have implemented and analyzed satellite protection techniques in space and predicted threats to satellites from celestial bodies. We have developed an algorithm where we have used deep learning concepts and astrometric calculations to gain greater accurate consequences. The outer space surveys have obtained a valuable understanding of the celestial bodies using the satellites observation process. Launching a single satellite costs between $10 million to $400 million so if a satellite gets damaged then it is a great threat to the economy and resources. Therefore, it is a necessity to detect an asteroid coming towards the satellite and deflect the satellite from its orbit.
Threats to satellites in space are first space weather, While space seems to appear an empty, vacuum void, the space environment is extremely complex. Our solar system is bombarded with cosmic rays and solar storm salvos of energetic particles, all of which can penetrate a satellite, wreck its electronics microscopically, and, in extreme cases, render it useless. Secondly, Congestion in space, Currently, nearly 3,000 operational satellites are orbiting Earth, and this number is increasing due to the abundance of small satellite launch opportunities, mainly into low-Earth orbit (LEO). The probability of accidents is increasingly growing. A single small collision could send thousands of BB-sized pellets hurtling through space at thousands of miles per hour in different directions. As space becomes more congested and contested, protecting our properties becomes increasingly necessary.
The interest in the detection and tracking of Asteroids has grown dramatically over the previous few decades. At the same time, technology has also been increasingly growing, enabling well-known surveys and amateur mini-surveys to decorate their solutions to attain better consequences. In light of this situation, we are making an attempt to use more reliable technology for detecting asteroids and predicting collision threats for satellites. In our proposed system the satellite will identify any risk in its immediate vicinity. Because the scope for detection from Earth via a telescope or sequential photos is so large, that the satellite protection process or the discovery of hazards near any satellite may be delayed. However, the ability of a satellite to identify itself for a specific surveillance region is limited. As a result, the model’s accuracy improves.
There are numerous attempts, studies, researches and software to locate asteroids. The previous researchers have done great work by providing tools that bring together all of the necessary components and simplify the analysis process for end users, the paper aims to enable effective anomaly analysis and attribution. A comprehensive satellite anomaly attribution tool is being developed by authors. While there is little that can be done to prevent space radiation anomalies once they are in orbit. The models and tools presented here will provide real-time data for making confident decisions in the event that problems arise and will help to ensure that satellite operations remain stable in the future. Some research papers have discussed several current aspects of radiation belt research, focusing on those most relevant to space weather applications where the term “space weather” refers to both natural and anthropogenic conditions. The most well-documented example is a large interplanetary shock event that produced an extremely intense belt of MeV electrons at L-shells where a HANE belt would be expected. The event was observed by the USAF/NASA Combined Release and Radiation Effects (CRRES) satellite in March 1991 and was still very strong when the mission ended in October of the same year. Another research work also describes operational concept, and system design for the Space Object Identification Satellite (SOISat’s), a Canadian Space Situational Awareness system. Simulation scenarios are included to validate SOISat’s performance in detecting and tracking resident space objects of interest.
In our proposed system the model will take input of asteroid images. A python module called Pillow is used to modify the scale of each image. These pre-processed images will be used for the next step of training the dataset. Training of the model is done using YOLO for real time detection, which requires labelled images obtained from OiDv4 toolkit which uses python library labelIng. Models can be trained end-to-end to improve accuracy. YOLO is more generalized. It outperforms other methods when generalizing from natural images to other domains like artwork. After successful training, the model will be able to detect real objects based on the dataset provided to the model. As our object is successfully detected we would move forward towards astrometric calculations which would help in determining the threat towards the satellite.
Proceeding with astrometric calculation, direction of the real-time detected asteroid is an important factor. The model is taking two pictures of the asteroid and comparing the two photos to see the position or size of the asteroid with respect to the satellite’s position. Next is the distance calculation between satellite and asteroid. Now for the calculation of the distance factor, which is the distance of an asteroid from the satellite, we have used concepts of triangles. Triangle similarity is used to determine the distance between our camera and the detected asteroid. We also need the speed of the asteroid so that using distance and speed we can find the time of the collision. Once all these calculations are done then we are calculating the size of the asteroid so that we can predict how much area is going to be destroyed and then can estimate the risk of collision between satellite and asteroid. After all the astrometric parameters are calculated the value of each will be shared on the earth centre where depending on values the risks are predicted and can decide the planetary defence technique for the protection of satellites. The main outcome of this system is to detect any threat to a satellite from an asteroid approaching it.
Date of Conference: September 14-17, 2021
Track: SSA/SDA