Spacecraft Attitude Takeover Control Via Multiple Cells

As a novel spacecraft architecture,cellularized satellites possess of great advantages of fast response,low cost,and convenient for on-orbit maintenance,upgrade,and reconstruction.This provides a new option for increasingly complex and diversified space missions.The cellularized satellites are constructed by some standardized and modularized satellite components,namely,cells,which are obtained by splitting the traditional satellite according to structural and functional characteristics.To meet the different requirements of specific mission,cells with corresponding functions can be selected to form a new satellite system.The application scenarios of cellularized satellite include replacing a specific part of the target satellite with the specific functional cell;and taking over the attitude control system of the target satellite by installing multiple cells.Based on the above background analysis,the problem of the design of attitude control law and torque allocation about multiple-cell satellites is studied.The main research contents are as follows:The attitude control system in attitude takeover mission consists of sensor cells,central controller cells and actuator cells.Considering the cells are connected through wireless networks,the communication capability of the cells is limited,and the energy consuming is with a high level.To reduce the communication pressure and energy consumption on the cell network,an event-triggered attitude tracking control law is proposed.The attitude takeover control system based on multi-cell structure faces the following problems: 1)Due to the distributed torque allocation strategy adopted by the flywheel cells,the central controller cell cannot obtain the total angular momentum of the flywheel cells,which makes the total angular momentum play a disturbing role.2)Due to the attachment,reconstruction and update of cells after attitude taking over,the moment of inertia of the combined system has great uncertainty.3)The disturbance caused by geomagnetism.In order to reduce the influence of the above uncertainties,an adaptive event-triggered control method based on neural network is proposed.The proposed method has the advantage of reducing the communication frequency between cells: only when the attitude tracking error exceeds the set threshold will the sensor cells be triggered to transmit measurement information.At the trigger moment,the controller cells will receive the measurement information and calculate the desired torque and transmit it to the actuator cell network.Besides,due to the used flywheel cells with a large number and a certain probability of failure,the output torque may have some errors.To deal with this issue,an adaptive fault-tolerant attitude tracking control method based on event-triggered is proposed,which can maintain a certain control ability even when some flywheel cells fail.Considering that the number of actuator cells is large,and the spatial layout is scattered,if the central controller cell communicates directly with all the actuator cells and allocates torque at the same time,the central controller cell needs to be equipped with a high-frequency communication system.However,this will bring high energy consumption.In addition,the central controller cell has other tasks besides attitude control.Therefore,if the central controller cell can be separated from the torque allocation task,the communication pressure of the central controller cell can be reduced and the energy burden can also be reduced.To this end,a distributed torque allocation method is proposed,which means that the actuator cells determine their own torque output based on the information of themselves and theirs neighboring cells,without requiring the central controller cell to perform torque allocation.The distributed torque allocation algorithm can achieve two targets.On the one hand,it ensures that the sum of the output torque of the flywheel cells is equal to the expected torque,that is,the allocation errors are within the allowable range.On the other hand,the angular momentum of flywheel cell reaches the same level with guaranteeing the allocation accuracy.The former is the main target of torque allocation,and the latter is an important way to manage the angular momentum of the flywheel cells,through which the angular momentum can be exchanged between the flywheel cells.In order to realize the torque allocation among the actuator cells in a distributed way,the torque allocation problem is transformed into a distributed convex optimization problem.The balance of angular momentum is formulated as the objective function of the convex optimization problem.The objective of making the output torque equal to the desired torque is formulated as equality constraints of the optimization problem.The torque saturations of the flywheel cells are formulated as inequality constraints of the optimization problem.Then,based on the consistency theory,a distributed torque allocation method is proposed by solving the convex optimization problem in a distributed way.In addition,since the flywheel cells tend to angular momentum saturation due to the absorption of constant disturbance in practical applications,a distributed flywheel angular momentum unloading strategy is proposed based on the angular momentum balance capability of the distributed torque allocation algorithm.This strategy achieves the goal of unloading the angular momentum of all flywheel cells by unloading the angular momentum of a specific individual flywheel cell.