Active Control Of Shape Accuracy Of The Large Solid Reflector Antenna

Due to its high shape accuracy,the large-scale solid reflector deployable antenna is widely used in the fields of space exploration,aerospace,and satellite communications.Because the gain of an antenna is closely related to the shape accuracy of its reflecting surface,and under the action of thermal load,the reflecting surface produces thermal deformation,which increases the surface error of the antenna.Therefore,it is very important to actively control the shape accuracy of the antenna's reflecting surface.This paper takes the solid reflector antenna as the research object,and studies the temperature field distribution,thermal deformation of the reflector,static control of the reflector,and dynamic control of the reflector during the operation of the antenna.The main research contents are as follows:(1)A finite element model of the antenna in orbit is established according to the space thermal environment and the heat transfer method of the antenna structure,and the satellite is set in the software according to the actual working conditions orbit around the earth,solver parameters,sun parameters,etc.The simulation results show that there is a temperature gradient of 280 K on the antenna reflecting surface.Under such working conditions,the reflecting surface is prone to large thermal deformation,which affects its working performance.(2)Establish the finite element model of the antenna structure,and apply the maximum temperature gradient of the temperature field solved as the thermal load to the reflecting surface to obtain the thermal deformation of the reflecting surface,and use the thermal analogy method to drive the actuator voltage equivalent to the nodal temperature load,the static control model of the reflecting surface is established from the perspective of statics,and the optimal control problem is transformed into solving the mathematical extreme value problem.Under the constraint of inequality,the sequential quadratic programming method is used to find the work The optimal drive voltage of the actuator.Numerical simulation results show that the profile error before adjustment is 94.1?m,the profile error after adjustment is 29.5?m,and the profile accuracy is improved by 68.7%.(3)Modal analysis was performed on the antenna reflecting surface.The results of the modal analysis showed that the mode of the reflecting surface did not have dense frequency characteristics.In order to reduce the calculation scale and achieve real-time control,the model was rationalized by the modal truncation method.Simplify,establish the dynamic model of the antenna system,transform the dynamic equation into the state equation in the state space,introduce the model predictive control method to actively control the reflector profile,and use numerical simulation to check the effectiveness of the model predictive control.The results show that the initial profile error of the reflective surface is about 700?m.After active control,the profile accuracy of the reflective surface is reduced to about 0.08?m in 0.5s,and the initial node displacement of the reflective surface is between-3mm and 0.5mm.When the control time is 0.5s,the error of each node of the reflecting surface is-0.0015 mm to 0.0025 mm,and the amplitude of the node error of the final reflecting surface is controlled within 0.002 mm.(4)Taking into account the antenna reflector model error and the external interference in the control process,the RBF neural network adaptive control is used to actively control the reflector profile accuracy.Due to the existence of model errors and external interference,error compensation needs to be considered in the control process.RBF neural network has the characteristic of approximating nonlinear functions.This means that the RBF neural network is used to estimate the error during the control process to achieve the accuracy of the antenna reflector surface.Dynamic adjustment.The simulation example compares MPC control with RBF neural network adaptive control.The comparison results show that when there is a large load interference from the outside,the profile error of the reflector controlled by MPC fluctuates greatly,and the control is not easy to reach a stable state.RBF neural network adaptive control has relatively small fluctuations and good robustness.