The increased collision risk for space missions posed by jettisoned engines and disintegrated spacecraft has Optimized BP neural network flow chart. Source: T. Ma et al.spurred the establishment of scores of space debris monitoring stations, most deploying laser ranging technology. The accuracy of these systems is currently constrained by an inability to precisely point telescopes at the position of orbital junk because of its small size and lack of surficial reflection prism. Now a new set of algorithms devised by researchers in China for laser ranging telescopes has improved the success rate of space debris detection.

Laser ranging technology uses laser reflection from objects to measure their distance, but the echo signal reflected from the surface of space debris is very weak, reducing accuracy. To address this limitation, the researchers trained a back propagation neural network to recognize space debris using two correcting algorithms. The Genetic Algorithm and Levenberg-Marquardt optimized the neural network's thresholds for recognition of space debris, ensuring the network was not too sensitive and could focus on localized areas of space.

The research team from Liaoning Technical University and the Chinese Academy of Surveying and Mapping used observation data for 95 stars to solve algorithm coefficients from each method, and the accuracy of detecting 22 other stars was assessed. A telescope corrected by the optimized neural network model was used to track space debris, demonstrating a pointing accuracy increase to nine times in the azimuth and three times in the pitch compared to that before correction. The accuracy of the model described in the Journal of Laser Applications is superior to that of three traditional telescope models and also overcomes the shortcomings of slow convergent speed.