Research On Identification Of Inertial Parameters Of Non-coopeative Target With Touch

As the development of space technology, the orbital operation technologies ofnon-cooperative targets is getting more and more attention in the world. The orbitaloperation technology can completes many space tasks,such as on-orbit assembly, on-orbit repairing, on-orbit maintaining, cleaning space junk, capturing or destroying theenemy’s military satellites. So, this technology has a high civil and military value.The measurement technology is the foundation of orbital operation technology, andvision-based measurement technique has been relatively mature, but the visualmeasurement techniques can only measure kinematic parameters, and does notidentify the inertia parameters, and inertial parameters measurement is still in theexploratory stage. The accurate inertial parameters are directly related to theimplementation of the orbital operation. At present, our study of inertial parametersidentification of non-cooperative target is still leaving behind the developedcountries. Therefore, the research on inertial parameters identification of non-cooperative target is very urgent and of great significance.Through the comprehensive analysis of spacecraft inertial parametersidentification methods, a inertial parameters identification method based on theconservation of momentum is adopted. The method completes the parametersidentification, by space robot rotating the antenna to motivate non-cooperative target,with the kinematic data information of space robot acquired by satellite-bornesensors and the kinematic data information of target got by the visual measurementtechniques.For the analysis of the identification process, this paper determines the simplephysical model, and builds kinematics and dynamics model of space robot and target,which lays the theoretical foundation of follow-up study.According to the analysis of the identification process, the linear regressionequation of inertial parameter identification is deduced. This paper makes the erroranalysis of the linear regression problem. Based on the research status, this articleuses the least square method and recursive least squares algorithm to identify inertialparameters from the linear regression equation, and analyzes the influence of inputdata. Based on the two kinds of recognition algorithm, this paper presents leastsquare algorithm scheme and the recursive least squares algorithm scheme based onthe least squares. The former identification accuracy is high, but too muchcomputational resources were wasted, and the latter is slightly low, but of highcomputational efficiency.In order to further improving the identification process, this article studys thekey phrase of the identification process-collision, and collision system dynamics model is established, ensuring antenna give target the controlled and incentivecollision. Calculating the collisions position between antenna and edge of target, thisarticle builds the collision contact model based on the equivalent spring dampingmodel and the coulomb friction model, and then models a dynamic model of collisionexcitation, and resolving model with the fourth order Runge-Kutta. Compared withthe simulation results of ADAMS, this article verify the accuracy of the systemdynamics model. Finally，the collision excitation effect influenced by error ofinertial parameter estimates was discussed.