| As human begins exploring the outer space,the demand for super-large-aperture space remote sensors are becoming increasingly urgent.However,due to the impact of the launch size of the spacecraft,the integral space remote sensor cannot meet the caliber requirement.The split type super-large-aperture space remote sensor has become an effective solution to such problem,and the split space assembly system is one of the most competitive technical solutions.In this technology,there is a need to perfect the technology related to the space flying robot,especially the autonomous navigation ability and motion control technology for robot.The improvement of these technologies can help the space flying robot to accomplish space tasks autonomously.In addition,the corresponding ground simulation experiments should also be carried out.The ground simulation experiment can help to verify the validity and reliability of the related technologies and accelerate the practical application of the space flying robot.Based on actual engineering and predicted requirements,this paper studies the motion control technology and ground simulation experiment technology for the space flying robot.The details are as follows:The kinematics and dynamics models of space flying robot were established,including the transformation of coordinate system,the relationship between kinematics and dynamics,and the arrangement of nozzles.Although these theoretical models have been applied,they are the basis for the implementation of robot trajectory planning and control algorithm,and are still necessary for analysis.Based on the mathematical model of a single robot,the nozzle reconstruction algorithm was studied when multiple robots work together,and the three distribution modes were designed: the standard mode,the energy-saving mode and the differential mode.In analyzing the reconstruction algorithm,it was found that the differential mode can be a good solution to the minimum pulse width limitation problem.This mode can be used not only in multibody system,but also in the control of a single space flying robot.The pulse modulation of a single nozzle is theoretically analyzed,and the pulse width modulation(PWM)method under the standard condition,the condition with a minimum pulse width limitation and the condition with a pulse width saturation problem are given,respectively.The experimental results showed that the established nozzle layout and designed PWM method are effective,and the nozzle control effect combined with the differential mode is better than the traditional nozzle control effect.Based on the mathematical theoretical model of the space flight robot,the algorithm of the space flying robot trajectory planning was studied,and a multi-window dynamic window method based on fuzzy rules and virtual target points was proposed.First,the traditional trajectory planning algorithm,including the trajectory planning based on the variational method and the trajectory planning based on the numerical method of the spline-based algorithm,was analyzed.The analysis showed that although such methods can plan the trajectory of the space flying robot in a known static environment,there are less capable of handling dynamic environments.Considering that the space flying robot are also subject to the limitations of the movement and the working environment,the dynamic window approach was introduced in this paper,and a dual dynamic window and virtual target are established to solve the minimum problem caused by the planning algorithm due to low speed or U-shaped obstacles.Fuzzy rules were used to adjust the weight of each parameter of the evaluation function to improve the adaptability of the robot to the dynamic environment.The simulation results showed that the algorithm can be applied to the cooperative work of multiple space robots;the algorithm can effectively avoid the minimum problem;the algorithm can effectively avoid dynamic obstacles.In order to ensure that the space flying robot can accurately track the planned trajectory,the motion control technology for the space flying robot was studied.In the study of such technologies,it is inevitable to encounter problems such as nonlinearity,time-varying parameters and coupling.Therefore,this paper studied the sliding mode control algorithm(SMC)and proposed a double closed-loop fuzzy sliding mode control algorithm with boundary layer.Fuzzy rules were used to reduce the control gain when approaching the sliding surface to smooth the control signal.The fractional boundary layer and dynamic boundary layer were designed respectively to solve the high frequency chattering in the SMC.The simulation and experimental results showed that the control algorithm has less chattering than other similar algorithms,and better control effect than the traditional PID control algorithm.To verify the validity and reliability of the above algorithms,a flying robot simulator based on cold gas propulsion and capable of free floating was designed.The simulator has the advantages of being able to load a variety of tools,having large carrying capacity and long duration.In the structural design of the simulator,modular design was adopted.The different function areas were divided into different modules to realize the unit assembly of the whole structure.In order to meet the load-bearing requirements of the entire simulation device,a self-planar air bearing device was designed to enable the entire device to have a load-bearing capacity of 800 Kg.Focusing on the performance of the simulator propulsion system,combined with the nozzle thrust experiment and the bearing capacity experiment,the rated air pressure of the ground experiment system was determined to be 0.4MPa,and the overall running time could reach 30 min.The localization algorithm of the ground simulator was also studied,and a localization algorithm based on the laser radar and gyroscope was proposed.Compared with other localization methods,this algorithm is simple,economical and highly applicable.The experimental results showed that the developed ground simulator system can meet the experimental requirements well and the proposed two-dimensional localization algorithm is feasible and reliable. |
PDF Full Text Request