Spatial non-cooperative target angular velocity measurement method based on convolutional neural network

Abstract

The invention discloses a spatial non-cooperative target angular velocity measurement method based on a convolutional neural network, and belongs to the field of spaceflight. The spatial non-cooperative target angular velocity measurement method comprises the steps of: generating training data by using a simulation method, and training a convolutional neural network to obtain a trained convolutional neural network; and preprocessing original distance data measured by a laser ranging radar, and inputting the preprocessed original distance data into the trained convolutional neural network to obtain an angular velocity of a non-cooperative target. The specific process of preprocessing is as follows: rearranging all elements in an original distance data matrix, and for the elements in the ithrow and the jth column, if the elements are moved to the (2i-1)th row, otherwise, moving them to the second (Ni+1)th row; and if the elements are moved to the (2j-1) th column, otherwise, moving theelements to the second (Mj+1) th column. The spatial non-cooperative target angular velocity measurement method greatly improves the recognition capability of the convolutional neural network, and improves the measurement and calculation precision of the angular velocity of the non-cooperative target.

Description

technical field [0001] The invention belongs to the field of aerospace, in particular to a method for measuring the angular velocity of a space non-cooperative target based on a convolutional neural network. Background technique [0002] With the expansion of the scale of human space exploration and development, the demand for on-orbit service for space non-cooperative targets is also increasing. Space non-cooperative targets mainly include on-orbit faulty or ineffective satellites and various space debris, and their on-orbit service has two important meanings: first, for faulty or ineffective satellites, repair them, refuel them, etc. On-orbit service can greatly prolong its on-orbit lifespan and significantly reduce the cost of space exploration and development missions; second, the removal or recovery of space debris occupying important orbits can reduce the risk of satellite-debris collisions and improve the safety of the space environment. safety. [0003] The key to ...

AI Technical Summary

Problems solved by technology

However, convolutional neural networks are usually used to process image data. When they are directly used to process lidar data, they often cannot accurately extract the key features of the data, and are prone to underfitting or overfitting. Therefore, difficult to apply in practice

Images