Research On Optimization Design And Identification Method Of Mass Property For Three-axis Air-bearing Simulator

Three-axis air-bearing simulator,as a satellite's full physical simulation test platform,can be used to simulate the micro-gravity,low-friction space environment and support full-star in-orbit attitude dynamic tests on the ground.To ensure the success of the tests,it is necessary to minimize the deviation of the center of system's mass from the centroid of the air floatation ball bearing as much as possible.And the moment of inertia for simulator must be similar to that of a real satellite.Therefore,the optimization of the layout for the simulator' equipment must be done at the initial design stage.Considering that some components' moment of inertia are unknown and cannot be modeled,the actual mass property is different from the theoretical value of design.In order to determine the quantitative assessment of the whole simulator,it is necessary to identify the mass property.So it's necessary to have a research on optimization design method and identification algorithm of mass property for the simulator.The main contents are shown as follows.Firstly,a load layout optimization design method based on particle swarm optimization(PSO)algorithm is proposed for the optimization design of the simulator mass property.Taking load center position and load rotation Angle as design variables,and taking the center position of mass and the moment of inertia as the optimization criterion,a optimization model of layout for the equipment is established.Then a geometric noninterference algorithm between loads is presented.By using linear weighted sum method,the multi-objective optimization problem is transformed into single objective optimization problem.The layout optimization design method is proved the works well by mathematical simulation.Secondly,in order to identify the mass property of the simulator,the attitude dynamics model of simulator is established.One model is based on 3D pendulum,while another is simplified.It is found that there is very little difference between the two models in the case of a small center of mass deviation.Thirdly,two joint parameter identification methods for the moment of inertia and the center of mass position of simulator are proposed.One is based on the attitude dynamics equation and combined with tracking differentiator,which is called TD-RLS.While another is based on attitude dynamics equation in integral form,which is called IRLS.Through mathematical simulation,it is found that the IRLS method is superior to TD-RLS method in the aspect of identification accuracy and convergence speed.But both of these methods have very low accuracy when identify some small parameters such as the product of inertia and the position of the center of mass.At last,in order to solve the problem of low identification accuracy of some parameters with small order of magnitudes,a concurrent recursive algorithm for identification of the mass property of simulator is proposed.In order to find the difference between the joint identification method and the concurrent recursive method,the singular value decomposition and condition number of the measure matrix are used to analyze the identifiability and identifiability degree.It is found that concurrent recursive algorithm can greatly reduce the condition number of observation matrix and improve the performance of observation matrix.The result shows that the concurrent recursive identification algorithm can greatly improve the identification accuracy of small parameters by mathematical simulation.