| Since human beings have engaged in space activities,a large number of spacecraft have been launched.Many spacecraft are damaged in scientific research mission or can't be recovered in time after completing the mission.They will rotate around the earth at a high speed in the form of space debris.Laser ablation is one of the most promising active debris removal technologies,which has many advantages such as wide range of operation,easy operation and reusability.The key of laser removing space debris is to make corresponding driving strategies for space debris of different shapes.As the current technology is not ripe,it is difficult to carry out corresponding research in the space environment.The main research centered on the ground-based simulation study about the laser-driven process of the space debris.The target is to finish the recognition algorithm about four typical-shaped targets of sphere,cube,cylinder and cone,and to establish the model of laser-driven to manipulate typical-shaped targets exactly.The fuzzy degree of the target images is evaluated with energy gradient method and the correctness of the algorithm is verified by simulation.The simulation result shows that with the increment of image fuzzy degree,the value of energy gradient decreases monotonously.The recognition algorithm of images with different fuzzy degree is studied.Firstly,the effect of common feature extraction algorithms on images with different fuzzy degree is compared.And the conclusion is obtained that with the increment of image fuzzy degree,the performance of the artificial feature operator is poor.Then,based on the principle of convolutional neural network,the structure of convolutional neural network is designed to recognize the target image.Finally,the correctness of the algorithms are verified by experiments.The experimental results show that when the ambient light is stable,the energy gradient method can accurately represents the fuzzy degree of the images and the convolutional neural network have a high recognition accuracy for images of different fuzzy degree.The pose transformation relation between the coordinate systems related to the process of laser driving targets is established.And the method about how to represent the attitude of the target is determined.According to the theory of laser ablation and reverse jetting,a motion model of four typical-shaped targets is established which is applicable to the case that the laser beams are focused into a small focal spot on the target surface,in order to calculate the increment of translational and rotational velocities when the targets are in the different attitudes.Compared with the model which is established in another reference,the correctness of the model established in this thesis is verified.The principle of how laser drive space debris is verified by the experiment of the laser ablation driven.According to the experimental results,the model of laser-driven about four typical-shaped targets is simulated.First,the process when laser irradiate the targets once is simulated and the displacement of the targets in all directions is obtained.Then,the process when laser irradiate the targets continuously is simulated and the translational and rotational velocities of the targets in all directions are obtained.The trend of the velocity of the targets is studied.The moving process of the targets and the influence of target attitude on motion state are analyzed.Finally,the trend of the velocity of the targets are studied by simulation when the initial attitude or the laser emission frequency is different.The results show that the smaller the angel between the laser incidence direction and the normal vector of the target surface in the initial condition is or the higher the laser emission frequency is,the faster the velocity of the targets along the laser incident vector direction is in the whole driving process,and the faster the angle between the direction of the target velocity and the direction of the laser incident vector decreases. |
PDF Full Text Request