The design scheme and in-orbit verification results of the high-precision star tracker is presented, which is used in the Shenzhou series of spacecraft in China. In the lens design, a front-placed diaphragm structure is adopted to meet the needs of miniaturization and long-term applications; In the overall structure design, a "frame combination wrap-around" integrated structure is adopted to focus on lightweight and stability; In the circuit design,radiation-resistant reinforced APS image sensors, processors and ASIC are applied to ensurance of high-sensitivity detection and high-reliable information processing throughout the service life; In the algorithm, clustering extraction, fast triangular recognition, curtain compensation and dynamic adjustment of exposure time, etc. are used to achieve key performance indicators such as precision, update rate, dynamic and capture; In addition, the thermal stability of the whole machine is improved to ensure that the performance of the star tracker can be maintained. The star tracker is successfully applied to a series of major space projects, including manned spaceflight, lunar exploration and Beidou-3 which have more than 200 units in orbit operation.

According to calibration problem of ten redundant laser gyroscopes ‘three-autonomy’ strapdown inertial measurement units, A calibration model and calibration scheme are deduced, which includes five-axis gyros and five-axis accelerometers. The ten-axis redundant inertial measurement unit is divided into two groups known as straight assemble X\Y\Z and oblique assemble S\T. A system level calibration method is adopted for the straight assemble, by establishing a state equation about the calibration error model, taking navigation speed error as the measurement, and estimating the calibration error parameters through Kalman filtering. Using the calibration results of the straight assemble as reference, separate calibration models are established for the accelerometer, gyros scale coefficient and gyro bias. Finally, the calibration of ten redundant laser strapdown inertial units is verified by combining the 19-position system-level calibration method with the discrete calibration method. The calibration results are correct and the characteristics of the calibration errors are analyzed.

The trajectory planning in task workspace that can occur singularities during a space free-floating manipulator operating is investigated. Based on the analysis and proof of the factors influencing the singularities, three profound characteristics and one measurement are presented, and then a smooth trajectory planning algorithm is designed by the robust pseudo-converse method. According to the SVD approach, the mechanism of the robust pseudo-converse method is presented, and the tracking errors in joint workspace and task workspace are estimated while avoiding singularities. On this basis, an improved robust pseudo-converse method for singularity avoidance is developed, which further reduces tracking errors. Numerical simulation results validate the correctness of theorical analysis and effectiveness of the proposed trajectory planning methods.

In response to the requirements of aerospace electromechanical servo systems for fast step response, no overshoot and high steady-state accuracy during tracking continuously changing signals, a composite control is involved on the basis of position-speed-current triple closed-loop feedback control with introduction of feedforward control. The tracking accuracy of continuously changing signals without reducing the dynamic performance and stability of the system is improved by using this control strategy. The design methods of closed-loop controller and feedforward controller are analyzed.Regarding the noise sensitivity of the feedforward function designed by invariance principle, a designed method based on error coefficient method is developed, and experimental verification is conducted. The experimental results show that the servo mechanism system can respond to step commands quickly without overshoot. The system has good tracking accuracy for continuously changing signals, and the system bandwidth has been expanded. The experimental results demonstrate the effectiveness and practical feasibility of the control method proposed in this paper.

The navigation and guidance challenges are researched for multi-event jettison resistance control probe during Mars aerocapture. A dynamic and measurement model for multi-event jettison drag modulation probe during Mars aerocapture is developed. Mars atmospheric density is modeled as a Gaussian process that is combined with unscented Schmidt-Kalman filtering to take into account the impact of atmospheric density uncertainties on the probe's system state and provides statistical information of state. A stochastic guidance strategy is employed, using unscented transform to propagate statistical information, which is incorporated into guidance target design and optimizes drag-skirt separation timing. Simulation results validate the approach's effectiveness by enhancing navigation precision and reducing required velocity increments for orbit entry. These advancements improve the success rate of Mars aerocapture and can serve as theory reference and technical supports for precise navigation and guidance in deep space exploration tasks.

Regarding meeting the task requirement of rapid missile launch, a fast firing data solving method based on neural network is proposed. Firstly, the mapping relationship among terminal velocity, altitude, velocity inclination and firing data is established, and the LM optimization neural network model is derived. Then, based on Bayesian regularization theory, the BP neural network structure is designed, and the optimized network structure which can meet the accuracy requirements is obtained. Finally, Newton iteration method is used to generate the database as training set which is used to train the neural network, so that the network model with optimized parameters is obtained, and the simulation is implemented. The theoretical and simulation results show that the rapid calculation of the firing data before shooting can be achieved by applying this method.

The integrated motion and control problem of attitude-orbit coupling for fluid-filled spacecraft in the presence of parameter uncertainties and external perturbations is focused in this paper. The liquid shaking is equivalent to a first-order spring-mass model, and a coupled dynamics model for the attitude-orbit of the liquid-filled spacecraft based on the dyadic quaternion is established, and the coupled dynamics equations for the attitude-orbit of the liquid-filled spacecraft and the equations of kinematics are deduced according to the laws of conservation of momentum and momentum moments. Regarding the external disturbances and model coupling characteristics, a preset performance terminal sliding mode control strategy based on an exponential form is proposed, which employs a disturbance observer to estimate the integrated disturbances and prove the stability of closed-loop systems by using Lyapunov theory. Finally, the numerical simulation results show that the fluid-filled spacecraft can realize accurate attitude-orbit control through the actions of the designed controller, which verify the effectiveness of the controller.

A dual-arm space robot finite time compliant capture strategy for non-cooperative targets is proposed. Firstly, the end-effector dynamic model of generalized coordinates presented by end-effector pose degree of freedom based on the free-floating space manipulator is established. On the basis of this, a global fast non-singular terminal sliding mode controller has been designed to achieve the finite-time trajectory tracking control of the end-effector. Furthermore, a three-fingered end-gripper mechanism has been introduced and a sub-system dynamic model consists of the end-effector and the target has been established. Based on this model, a composite compliant grasping control strategy has been designed, which is based on the compliant control theory to resolve the compliant grasping control of the end-effector towards the target. The simulation results show that the compliant capture of a tumbling target by a dual-arm space manipulator can be achieved by the proposed control strategy.

In response to the demand for automatic and rapid calculation of control strategies for maintaining the trajectory of a large number of satellites, research is implemented for the optimization design of satellite orbit control quantity calculation under the customized maintenance loop indicators. The optimization algorithm for satellite orbit control variables under customized maintenance loop indicators is studied. On the basis of analyzing the characteristics of the regression orbit, the calculation method of the reference benchmark for trajectory network prediction is optimized. Pseudo observation data is obtained by using historical precision ephemeris, and the average atmospheric damping coefficient is calculated as input for the trajectory network precision prediction model. An automatic calculation method for trajectory control quantity based on iterative correction is designed, which accelerates the search speed by reasonably selecting initial values, adjusting the coarse screening interval of control quantity and setting iteration termination thresholds. The simulation results show that this the automatic calculation of control variables for satellite trajectories that reaches the specified western boundary can be quickly completed by using this method which provides theoretical support for the automatic calculation of routine orbit control strategies for a large number of satellites in the future.

Regarding the planning of on-orbit refueling tasks for spacecraft in geosynchronous orbit (GEO), an on-orbit refueling strategy based on hierarchical refueling is proposed, which can effectively improve the refueling efficiency compared with the traditional “one-to-many” refueling method. In the proposed hierarchical on-orbit refueling strategy, a sub-service spacecraft (SSc) is defined to jointly accomplish the on-orbit refueling task with the primary service spacecraft (PSSc). On the basis of this, the spacecraft orbital transfer model based on the multi-turn Lambert orbital transfer, the relationship between orbital maneuvering speed increment and transfer time is obtained, and the fuel cost and transfer time of spacecraft orbital transfer are taken as the objective function under consideration of the constraints such as spacecraft load and mission time, and the spacecraft refueling sequence and orbital transfer time are optimally solved by using an improved genetic algorithm. Finally, the effectiveness of the theory is verified through numerical simulation.

Aiming at the problem of multi-coverage of hot regions,a design scheme of common ground track constellation with regression orbit attribute is applied,and a design method based on genetic algorithm(GA)for optimization of the orbital parameters of a single satellite and analytic method for solving the orbital parameters of a constellation is proposed in this paper. Firstly, a regression orbit model for muti-coverage of hot region is established.Secondly,by taking the optimal coverage effect of a hot region as the performance index,genetic algorithm(GA)is used to optimize and solve the orbit parameters of a single satellite.Finally,the orbit parameters of the constellation are solved by analytical method. The results of theoretical analysis and simulation totally show that the design method can achieve periodic multi-coverage of hot regions and higher time resolution ratio, which present fairly good engineering application value.

Under consideration of the difficulty in obtaining an accurate model of the flywheel and the limitation of computing power, a combined fault diagnosis method based on improved LSTM and fault tree is proposed. Firstly, the traditional grey wolf optimizer algorithm (GWO) is improved by population initialization, distance control parameters and α wolf position updates to achieve better convergence performance. Then, during the network training process, the improved GWO is introduced to optimize the hyper-parameter space, so the low efficiency of hyper-parameter selection caused by traditional manual adjustment method or grid search method is overcome; Further, due to considering the engineering practicality of fault tree analysis and the autonomy of neural network, a fault diagnosis framework that combines with those two ways is designed; Finally, a flywheel fault tree model is established and simulation experiments are conducted, which demonstrate the excellent convergence of the improved GWO and the effectiveness of the combined diagnosis algorithm for flywheel fault detection and recognition.