In this paper mainly starting from the development background and practical significance of space on orbit manufacturing technology, the development of on orbit manufacturing technology is classified, summarized and commented from the four aspects of on orbit 3D printing, on orbit welding, on orbit strip based plastic forming, and on orbit in situ manufacturing. A series of bottleneck problems encountered in the current development process are pointed out, and the future development trends are predicted. Combining with the successful experiences of foreign on orbit manufacturing technology, this article finally gives out the areas that need to be paid attention and the direction of efforts in the development of China’s on orbit manufacturing technology. It is expected to provide a useful reference for the layout of China’s space on orbit manufacturing technology.

In order to solve the problem of high maneuverability and difficult trajectory prediction of hypersonic glide vehicle (HGV), an intelligent trajectory prediction method of HGV based on ensemble empirical mode decomposition and attention long short term memory network is proposed by selecting the aerodynamic acceleration as the prediction parameter. Firstly, the maneuvering characteristics and the aerodynamic variation law of HGV are analyzed based on the six degree of freedom motion equation. The dynamic tracking model is established to estimate the aerodynamic acceleration in real time. Secondly, the estimated aerodynamic acceleration is decomposed and reconstructed by using ensemble empirical mode decomposition to weaken the influence of noise and avoid interference to the prediction model. Finally, the denoised aerodynamic acceleration data used to train the attention long short term memory network. Then the future aerodynamic acceleration data predicted and the future trajectory of HGV is reconstructed to achieve online trajectory prediction. The simulation results show that the method can effectively predict the maneuver trajectory of HGV with high prediction accuracy and good stability.

Based on fuzzy clustering and LSTM network, a data driven anomaly analysis method of launch vehicle engine oxygen turbopump data is proposed. Firstly, fuzzy clustering is used to pre classify the data samples with complex working conditions and incomplete labels to obtain complete labels and analyze the feature contribution, which lays a foundation for feature screening and training of LSTM network. The LSTM network is used to predict the data of the oxygen turbopump, and the average error between the predicted results and the original data is calculated. Then the threshold criterion calculated by the non parametric threshold calculation method is used to determine whether the turbopump is abnormal. Finally, the fault detection and alarm driven by the oxygen turbopump data are realized, and the accuracy is improved by 7% compared with the red line threshold detection method.

For manned lunar and deep space exploration missions in the future, the transfer design of low energy round trip circumlunar orbit and libration point in the Earth Moon system is studied, the effect of circumlunar orbit changes on key parameters such as flight time and mission fuel consumption are systematically analyzed under different three body orbits, and a search strategy for the initial value of round trip orbit design is presented. In order to solve the problem of initial value sensitivity of design variables, a differential correction algorithm is used to rapidly construct the initial transfer orbit by combining the invariant manifold with the characteristics of the libration point orbit. Considering multiple constraints such as near lunar and optimal fuel simultaneously, the round trip trajectory between the circumlunar orbit and libration point orbit is further studied through the multiple shooting method and quadratic programming algorithm, and the gradient formula of the constraint equations is deduced to improve the design efficiency. In order to analyze the characteristics of round trip orbits and verify the effectiveness of the design strategy, the relationship between the variation of parameters, such as amplitude of three body orbits, the inclination of different circumlunar orbits, and the transfer time and fuel consumption is studied. The design results have important reference significance for the design of round trip orbits and selection of parameters for the deployment of lunar exploration vehicles.

The development of four kinds of parabolic antenna structures is presented, including radial rib antenna structure, wrapped rib antenna structure, truss antenna structure and hoop antenna structure. A detailed comparison of these four structures is presented. Additionally, for the planar antenna structures, the features of wing antenna structures, truss antenna structures and membrane antenna structures are summarized. Finally, future trends of large, high packing factor and high surface accuracy are pointed out. And the technical difficulties which should be overcome before space applications of larger and higher precision SAR antennas are proposed.

The components and initial deployment of a space elevator system is briefly introduced in this work. On this basis, aiming at the complex dynamics and control problems existing both in the initial deployment and the follow up operation, the development status of dynamics and control of the space elevator system is analyzed and summarized from five aspects, including the dynamic modeling, stability, cable oscillation, cable oscillation caused by climber(s) motion and oscillation suppression, and initial deployment dynamics. Finally, further development of dynamics and control of the space elevator is summarized and prospected.

Aiming at the difficulties of a class of hypersonic flight vehicle (HFV) involving variable configurations, the engineering application requirements are analyzed, and the research progress of several typical nonlinear control methods is summarized. Firstly, the sources and characteristics of several common research models of hypersonic flight vehicle are summarized. Secondly, based on the actual engineering requirements, the control difficulties and needs for the control system capability of such vehicle are analyzed. Furthermore, the current status of several typical nonlinear control methods and intelligent control methods for hypersonic flight vehicle are summarized, and the frameworks of various control schemes are given. Finally, some problems and directions for further research are discussed for the HFV control with diversified mission forms and complicated environment in the future.

This paper proposes an integrated guidance and control algorithm based on deep reinforcement learning technique. Differently from the traditional integrated guidance and control algorithm and designing the guidance loop and control loop separately, the fin deflection command of proposed integrated guidance and control algorithm is given by the agent through the observation states of missile. The agent is generated by the deep reinforcement learning. To utilize the deep reinforcement learning technique in integrated guidance and control problem, we transfer the integrated guidance and control problem into a Markovian decision process that enables the application of reinforcement learning theory. A heuristic way is utilized to shape a proper reward function that has tradeoff between guidance accuracy, energy consumption and interception time. The state of the art deep deterministic policy gradient algorithm is utilized to learn an action policy that maps the observation states to a fin deflection command. Extensive empirical numerical simulations are performed to validate the effectiveness and robustness of proposed integrated guidance and control algorithm.

A deep learning based method is proposed to detect GEO objects from the low precision CCD images for the ESA “SpotGEO” competition. The Gaussian process regression and template matching method are adopted in the image data preprocessing step. According to the motion characteristics of GEO objects, the topological sweeping method is used as a preliminary step. To reduce the noise effect, an object filtering method is proposed. Two additional data filters are set before and after the topological sweeping respectively using the convolutional neural network. They significantly decrease the number of noise points and increase the detection accuracy. Results show that this method can reach a high detection accuracy of 98%, which is suitable for the sophisticated environment with light pollution and clouds covering.

Aiming at the challenges caused by the persistence, concurrency and energy consumption of probe actions, a plan repair method of deep space probe based on the expected state sequence is proposed. In this method, the expected state sequence is formed of the expected effect of the unfinished action and the expected precondition of the unexecuted action in the pre designed plan, according to the execution status of the action. The expected state sequence is an ordered set of states with mixed logic and energy, providing subgoals for plan repair and also transforming the plan repair problem into the state transition path searching problem. During the search, the plan repair strategy with energy supply priority is proposed, which separates the logic repair from energy repair to reduce the difficulty of solving the problem. And this method enables the probe to recover from plan failure autonomously. Finally, the effectiveness and rationality of the proposed method are verified through simulation by taking the Mars Orbiter as an example.

Centering on the structural mechanism technology in the on orbit construction of space optical telescope, this paper focuses on the basic concepts and technical characteristics of key technologies related to structure and mechanism, including topology optimization and module technology of space structure of large telescope, high precision deployment adjustment and locking technology of optical mechanical structure and mechanism, on orbit additive manufacturing technology of typical optical mechanical components. It also analyzes the technical connotation and advanced technical approaches of robot assisted autonomous precision assembly and control technology. Finally, the development trend of structure and mechanism technology in on orbit construction and some technical problems to be studied are prospected. This paper may provide a reference for evolution of space optical telescope.

This paper systematically reviews the types of parking orbits near the moon that can be used to deploy space station, and gives the definitions and main parameters of different types of orbits. The energy requirements, mission support, space environment and other characteristics of different types of orbits are analyzed through simulation calculation and comparison of domestic and foreign orbit research conclusions. The main considerations and basis for the selection of the deployment orbit for deep space Gateway are analyzed. Finally, this paper offers several suggestions on the application of various kinds of orbits, which can provide technical reference for the selection of the parking orbit for the future deployment of lunar space station.

In order to cope with the characteristics of short duration of Mars EDL process, large telemetry and telecontrol delay, large impact dynamic environment and shortage of high precision map, the entry capsule GNC system of Tianwen 1 probe is designed with autonomy and fault tolerance. The design of system level hot backup and heterogeneous hot backup of key components ensures the independent diagnosis and correction of component failure. The fault tolerance of the system is improved by the strategies of independent work mode conversion and the multi criterion backup of key work mode conversion. The autonomous impact resistance management of the components power supply as well as the autonomous fault diagnosis and correction strategy of touch down sensors ensure the effective response to the high dynamic conditions such as the initiation of initiating explosive device. Aiming at the combination of parachute backshell avoidance and terrain obstacle avoidance, a set of hazard avoidance system and its application strategy of laser imaging sensor and optical sensor are designed. Special experiments such as helicopter flight test are implemented to verify the system design. In actual flight, the entry capsule GNC system of Tianwen 1 probe successfully implemented the key actions such as unfolding the trim wing, completed the parachute backshell avoidance and terrain obstacle avoidance, and achieved the final landing accuracy of 3.1 km, which fully proved the correctness of the system design.

The basic principles of three modes for moving mass control are summarized, including the extra moment of inertia despin control, mass moment attitude control, and stability regulation. The key control problems that the moving mass high speed vehicles would face are then analyzed in detail with four aspects, i.e. dynamic modeling, guidance & control, actuator implementation, and system simulation & evaluation. Finally, the development trend and potential research themes are pointed out for the moving mass high speed vehicles.

The importance of the on orbit artificial intelligence technology for imaging satellite and the current research status are analyzed in this paper. Restricted computing power, high real time requirements, lack of large scale data sets, and special space environment are main challenges faced by the on orbit intelligence. Then the intelligent algorithms, hardware platforms, and optimization techniques are systematically summarized. As for the object detection, the remote sensing image data set, model architecture, real time and other aspects are introduced in detail. Finally, the future development of on orbit artificial intelligence technology for imaging satellite are studied and discussed.

In this paper, the acceleration technologies and basic theory of various types of electric propulsion, and the advantages and disadvantages of existing acceleration technologies are summarized. Then the development trend of acceleration technologies or the latest acceleration technologies both in China and abroad are introduced. On this basis, the next electric propulsion acceleration technology is prospected to provide reference for the research and development of new electric propulsion technology in China, and promote the further development of space propulsion technology.

Aiming at the problem of intercepting endo atmospheric high speed maneuvering targets, a deep reinforcement learning guidance law is proposed based on the twin delayed deep deterministic policy gradient(TD3) algorithm. It directly maps the engagement information to the commanded acceleration of the interceptor, which is an end to-end, model free guidance strategy. Firstly, the engagement kinematic model of both sides is described as a Markov decision process suitable for deep reinforcement learning algorithms. After that, a complete deep reinforcement learning guidance algorithm is constructed by reasonably designing the engagement scenarios, action space, state space and network structure required for algorithm training. The reward shaping and random initialization are introduced to construct a complete algorithm. The simulation results show that, compared with the proportional guidance and augmented proportional guidance laws, the proposed guidance strategy can reduce the requirement for mid course guidance while having smaller miss distances. It has good robustness and generalization ability, with less computational burden that makes it eligible to run on missile borne computers.

For a Mars aircraft exploration mission, a coaxial Mars rotorcraft is proposed. The structural parameters, such as the blade airfoil, planform, and twist angle, are optimized based on the computational fluid dynamics methods. The aerodynamic model of the rotor is established based on the blade element theory and momentum theory. The flight parameters such as rotor speed, rotor spacing, and blade installation angle are selected based on the numerical simulation methods, which prompt the design of the structure and control system of the prototype “MarsBird I”. A Martian atmosphere simulator and a combined device for gravity compensation and motion constraint are constructed to carry out the ground flight test of the Mars rotorcraft under a simulated Mars environment, which verifies the propulsion performance of the coaxial Mars rotorcraft. Besides, we prospect the research directions of the Mars rotorcraft technology. The research results provide an important reference for China’s Mars exploration project.

Taking the moving platform aircraft cluster as the research object, cluster cooperative positioning and navigation countermeasures technology is analyzed, as well as the state of the art. By refining technical difficulties, the development trend of the technology is put forward, providing new ideas for future construction of autonomous spatio temporal reference system and navigation countermeasures of aircraft cluster in China.

In order to improve the computational efficiency of trajectory replanning in ascending flight of launch vehicles experiencing thrust drop fault, an online trajectory replanning method based on the intelligent decision making is proposed by transforming the original optimization problem into two independent problems, i.e. the intelligent decision making of the optimal rescue orbit and the optimization problem to achieve minimum fuel consumption (MFC). The “fault state rescue orbit” sample dataset is generated by offline solving the trajectory replanning problem with all kinds of different thrust drop fault states. Based on the sample dataset, the decision making model of the optimal rescue orbit is established by training a radial basis function neural network. The rescue orbit decision model is online used in actual flight to decide the optimal rescue orbit elements corresponding to the current thrust drop fault state. Subsequently, the MFC problem using the rescue orbit as the target orbit is solved. And so the optimal rescue trajectory of the launch vehicle under thrust drop failure is obtained online. The numerical simulations show that the proposed method improves the computational efficiency by more than two orders of magnitude and gives the same optimal rescue trajectory, compared with solving the original trajectory replanning problem directly.