Abstract:With the development of on-orbit service technology and the control of spacecraft launch operating costs， the spacecraft on-orbit service mode will develop from a “one-to-one” service mode to the “one-to-many” and “many-to-many” service modes. In the “many-to-many” service mode， the mission allocation and planning for service spacecrafts will become particularly critical. Therefore， in this paper， the on-orbit refueling mission planning for multiple geostationary earth orbit (GEO) satellites with multiple service spacecrafts is studied. First， a many-to-many on orbit refueling mission planning model is established by considering many kinds of constraints， e.g.， service spacecraft capacity constraints and service path constraints， and taking minimizing fuel consumption as the optimization index and the service order of each service spacecraft as the decision variable. Second， in view of the poor local search ability of the genetic algorithm (GA) and the defect that it is easy to fall into the local optima， combining the large neighborhood search (LNS) algorithm and the GA， a hybrid heuristic algorithm (LNS-GA) is designed to solve the mission planning problem. The designed algorithm uses the “destroy” and “repair” ideas in the LNS algorithm to perform further iterative search on the elite individuals in each generation of the GA， and thus its local search capability is enhanced. Finally， based on the specific simulation scenario， the effectiveness and superiority of the algorithm proposed in this paper are verified by comparing with a conventional genetic algorithm.

Abstract:To optimize the maneuvering strategy of long-range transfer and proximity for geosynchronous earth orbit (GEO) satellites， the single pulse Lambert transfer within time limit is analyzed with contour plots of Pork-chop. A mathematical model of multiple-pulse transfer and proximity is built to find the global optimal solution with the differential evolution (DE) algorithm， taking the minimum fuel as the optimization goal. The trajectory proximity is calculated under different pulse numbers and different time constraints through simulation. The results show that the results of triple-pulse transfer and proximity are relatively optimal.

Abstract:A method based on multi-agent deep reinforcement learning algorithm is proposed for the multiple satellites and non-cooperative targets in the orbital pursuit-evasion game， which has complex dynamics model， unknown maneuver information of non-cooperative targets， and challenge to coordinate effectively among satellites. Firstly， the game scenario is modeled， the reward function is reshaped and improved under the scenarios of minimum time， optimal fuel and collision avoidance. The multi-agent deep deterministic policy gradient algorithm is used for centralized training to obtain the optimal pursuit policy parameters of each pursuing satellite and evading satellite. Then the distributed execution enables multiple pursuing satellites and evading satellites to complete the pursuit-evasion game. The simulation results show that the method can complete the pursuit-evasion game of multiple satellites against non-cooperative targets and use the numerical advantage to effectively increase the success rate of pursuit and reduce the energy consumption in the pursuit process. Moreover， a series of intelligent game behaviors such as “interception”， “siege”， “infiltration”， and “capture” emerge， which are conducive to effectively achieve the game purpose.

Abstract:Using robot to assemble spatial components is an essential way to build ultra-large spacecraft. Available studies on space assembly mostly focus on small structures， without considering the effects of gravity gradient and orbit-attitude-structure coupling. In this paper， an orbit-attitude-structure coupled dynamic model is established for the space assembly system including space robot and flexible structure to study the effects of gravity gradient and orbit-attitude-structure coupling on the assembly process. First， the space robot is regarded as rigid， and the main structure and the structure to be assembled are regarded as flexible. Moreover， natural coordinates are used to model the rigid space robot， and absolute nodal coordinates are adopted to model the flexible structures. Then， the Hamilton equation of the system is derived with the consideration of the kinetic energy， the gravitational potential energy， and the elastic energy of the rigid and flexible bodies， taking the effects of gravitational force and gravity gradient into account. Finally， orbit-attitude-structure coupled dynamic simulations are conducted by trajectory planning and trajectory tracking control， and the dynamic responses of the system are studied， including the orbital motion， attitude motion， control results in both the joint space and the Cartesian space of the robot， and structural vibrations of the flexible structures. The results show that the effects of gravity gradient and orbit-attitude-structure coupling are vital to the control moment of the space robot and the assembling precision， which must be considered in the assembly process design and control system design.

Abstract:The autonomous orbit keeping method for satellite constellations by means of electric propulsion is studied. Nominal orbit recursion is carried out from the ground and injected regularly. The quasi-mean orbit parameters of the local satellite and the reference satellite are recursed， considering the non-spherical J2 perturbation of the earth， and intra-planar phase keeping is controlled with the single side limit cycle method. A satellite autonomous orbit control scheme is designed for a fixed solar wing satellite running in the solar-oriented mode for a long time， considering the constraints of the energy， program control task， and orbit eccentricity. The effectiveness of the algorithm is verified with a high precision orbit perturbation model.

Abstract:A trajectory optimization method of remote rendezvous for upper stage based on sequential convex optimization is proposed for solving the problem of remote rendezvous for upper stage with free terminal time. First， the remote rendezvous model for upper stage with free terminal time is established. The problem is transformed into a series of iterative convex optimization problems by means of constraint relaxation and linearization. Second， considering the characteristics of large thrust of the upper stage engine， the result of Lambert two-pulse orbit transfer is adopted as the initial iteration value in order to reduce the iteration times. Finally， the remote rendezvous trajectory and thrust sequence are achieved by using the primal-dual interior-point algorithm. The numerical simulation results show that the proposed method has a fast calculation speed and real-time performance， and can provide reference for remote rendezvous task of upper stage.

Abstract:Wired-contact data communication and power supply are often used between guided weapons and launchers. In order to meet the needs of rapid loading and anti-saturation attack in the future， more attention should be paid to wireless energy-carrying communication. The data transmission technology based on wireless coupling is widely used in current wireless energy-carrying communication. However， in the increasingly complex and strong electromagnetic environment， this technology is susceptible to interference and thus fails. In order to overcome this deficiency， in this paper， a wireless energy-carrying communication system (WECS) based on optical carrier is proposed， which adopts the optical carrier modulation， transmission， and demodulation technology. The WECS utilizes the characteristics of visible light and infrared light to carry out reliable wireless communication and energy transmission in a strong electromagnetic environment， which can be effectively applied to the wireless launch scene of weapons.

Abstract:The radio over free-space-optical (RoFSO) link uses a free-space-optical link to transmit radio-frequency (RF) signals through atmospheric channels. Owing to the effects of atmosphere turbulence and temperature， the time delay of the free-space-optical link suffers from timing jitter， which may degrade the performance of the RoFSO link. In this paper， the absolute time delay of a 1 km RoFSO link is precisely measured with a phase-derived method. The results are accord with the theoretical model， and the measured accuracy is higher than 0.1 ps. Owing to the use of electro-optic modulation and radio frequency (RF) phase discrimination， the measurement method is compatible with the RoFSO link， and the high accuracy can meet the requirements for a stable RoFSO link.

Abstract:This paper proposes a global non-singular terminal sliding mode control method based on double power reaching law to solve the fast response robust control problem in an electric steering system. First， we establish the mathematical model for the steering gear system， with the aerodynamic disturbance and friction modeled as bounded disturbances. Second， we construct the non-singular terminal sliding mode surface according to the system model， and design the controller of the steering gear by combining the double power reaching law. As a result， we can ensure the fixed time accessibility of the sliding mode surface take full advantage of the robustness and invariance of the sliding mode control method， and improve the response speed of the steering gear system. Finally， analyses on the Lyapunov stability and numerical simulations are carried out for the controller. Relevant results show that the new control method makes the system error converge to zero within a finite-time， and effectively improves the performance of the steering gear system.

Abstract:In order to solve the problem of hyper-spectral detection of sub-pixel aircrafts， an improved RX anomaly detection algorithm based on atmospheric absorption spectra is proposed to improve the detection capability and detection time efficiency of the sub-pixels. The basic principle of the improved algorithm is expounded， and the difference curve of the atmospheric absorption coefficient between the 10 km altitude and the ground is research. The detection model of the improved algorithm is established. A hyper-spectral simulation image of a signal-to-noise ratio of 10 dB is obtained with the visible-short wave infrared hyper-spectral data of passenger aircrafts and the sea water hyper-spectral data by simulation. The classical RX algorithm and the improved RX algorithm based on 4 atmospheric absorption bands are used to detect 150 spectrum simulation images， and the detection capability and receiver operating characteristic (ROC) curves of the sub-pixels are obtained with the classical and improved algorithms. The results show that improved RX algorithm can effectively improve the detection capability of the sub-pixels and shorten detection time.

Abstract:In order to solve the problem of catalogue power supply in Mars exploration missions， the planning of taking liquid metal Rankine cycle as the thermal power conversion module of Mars catalogue base is developed. The Simulink modular modeling method is used to program the liquid metal Rankine cycle model. According to the coupling standard of each component of the space Rankine cycle， the simulation models for the system components and overall control are established. The fluctuation of the cycle output power under variable operating conditions is studied. The results show that the final output power of the Rankine cycle dynamic thermoelectric conversion simulation model can reach more than 1 MW under various working conditions， and the thermoelectric conversion efficiency can reach 27.3% under the rated working conditions， which fully meets the thermal power requirements of the system under the background of Mars detection. The Simulink modular modeling method can provide guidance for space nuclear power thermoelectric conversion simulation technology， and the relevant research results can provide valuable reference for China's Mars exploration plan in the future.

Abstract:Barrage jamming will submerge the target signal and make it unable to be detected by radar. In the presence of strong jamming， radar systems may be paralyzed directly. Therefore， it is of great significance to study the methods of suppressing barrage jamming. In order to suppress barrage jamming， an optimization problem model is built. The discrete phase coding sequence is used as the transmitting signal， the minimum range sidelobe of the transmitting signal is used as the criterion， and the fixed output level of the barrage jamming processed by the matched filter and the discrete phase coding are served as the constraints. The alternating direction method of multipliers (ADMM) nested quasi Newton method， called as ADMM-BFGS， is introduced to solve the problem. Moreover， in terms of the problems of complexity and long time-consumption existing in the ADMM-BFGS algorithm， a composite algorithm combining the ADMM with power method-like iterations (PMLI)， called as ADMM-PMLI， is proposed. The simulation results show that the ADMM-PMLI algorithm can greatly reduce the calculation time compared with the ADMM-BFGS algorithm， and both the ADMM-PMLI and ADMM-BFGS algorithms can ensure the detection performance of the radar and improve its ability to combat barrage jamming.

Abstract:Due to the large number of images， accurate and efficient target detection is the key step to enhance the automation of the shooting-range photometric image processing. Aiming at the problem of poor adaptability of traditional target detection algorithms due to multiple low-altitude target images and target types and changes in target characteristics， this paper proposes an automatic low-altitude target detection method based on dual-attribute classification deep learning. The method is based on YOLO V3， a deep learning target detection framework， and improves the single-attribute classification in the output layer of the network to dual-attribute classification based on the dual-attribute features of luminance and shape of the low-altitude target;achieves automatic sample annotation based on target region growth， and confirms the detection results using sequential image target constraints. The actual image training and detection results in the low-altitude scenario of the range show that the initial detection success rate of the method is higher than 90%， and 99% detection success rate and 62% average localization accuracy are achieved after post-processing.

Abstract:With the rapid development of Hall thrusters， iodine fueled Hall thrusters have been paid more and more attention from researchers. In order to deeply understand the internal procedure in the discharge channels of iodine fueled Hall thrusters and provide scientific evidences for optimizing the performances of thrusters and expanding their space applications， a two-dimensional particle-in-cell/direct-simulation-Monte-Carlo/Monte-Carlo-collision (PIC/DSMC/MCC) hybrid model is established with sheath and secondary electron emission. According to the properties of iodine， dissociation and ionization are added. Under the condition of constant channel wall temperature， numerical simulations are carried out for the internal procedure in the discharge channel of a 200 W iodine fueled Hall thruster. The multi-field coupling characteristics of plasma and the interactions between the channel walls and the plasma in the discharge channel are studied. The ion number density， ion axis velocity， and electronic temperature distribution are obtained to study the plasma actions and analyze the parameters of plasma. Compared with xenon， iodine fueled Hall thrusters have a dissociating region with a width of about 2 mm， which is located behind the near-anode region and before the ionization region.

Abstract:a four-point suspension method of hoop flexible structure is proposed， and modal test is carried out. Before the ground test of hoop flexible structure， the three-dimensional model of suspension device is designed for mechanical analysis. Firstly， the static balance equation is established in consideration of the deformation coordination condition， and the input voltage and output ampere force of multiple suspension motors are solved. Then， the natural frequency and static deformation of the structure are calculated by the finite element analysis. Finally， the four-point suspension device of scale model is fabricated to test. The results of suspension test show that the voltage of suspension motor is consistent with the value of theoretical analysis. The modal test results show that the natural frequencies before and after suspension are close， which proves the accuracy of the suspension device. In a word， it provides theoretical basis and experimental support for the design of multi-point suspension device of spacecraft flexible structure.

Abstract:A pre-swirl turbine disk structure is proposed to solve the problems of excessive temperature and large temperature gradient of the turbine disk surface of space nuclear power spacecraft. The surface heat transfer characteristics of turbine disks with or without the pre-swirl cooling structure are analyzed by numerical simulation， and the influence of the rotational Reynolds number on the heat transfer effect is studied. The results show that the pre-swirl structure can effectively improve the convective heat transfer coefficient of the helium-xenon cooling air flow and the turbine disk， and reduce the maximum surface temperature and temperature gradient. The maximum temperature with the pre-swirl structure is 63.1 K lower than that without the pre-swirl structure. With the increase in the rotational Reynolds number， the maximum temperature of the turbine disk surface decreases， while the average heat transfer coefficient of the turbine disk surface increases. The average heat transfer coefficient with the pre-swirl structure is 13.4% higher than that without the pre-swirl structure.

Abstract:To address the time-of-arrival (TOA) based localization problem when the low and high orbit satellites are not synchronized， a non-convex weighted least squares (WLS) localization equation is proposed， which in view of both the target position coordinates and the unknown transmission time. A transformation measurement model and the semi-definite relaxation (SDR) algorithm are used to relax the WLS problem into a convex semi-definite programming (SDP)， which has been proved to be always tight， and thus the optimal solution of the WLS problem can always be obtained. Furthermore， the proposed method is extended to a moving target localization scenario， where the target velocity is assumed to be constant in a short observation period. The proposed algorithm can effectively solve the problem of deteriorating satellite localization performance under the condition of clock synchronization error. The simulation results show that the localization accuracy of the proposed method can reach the Cramer-Rao lower bound accuracy when the noise is not very large.

Abstract:In order to meet the requirements of the form and position accuracy measurement and the deformation control in the assembly process of launch vehicle structure cabin， the accurate and efficient measurement technology for super-large cabin is studied. First， the advantages and disadvantages of different measurement technology schemes are compared and analyzed. Second， based on the photogrammetry technology principle， the process of photogrammetry for super-large cabin is established. With the process， the point cloud data are collected， and the product model is reconstructed based on the collected point cloud data. Third， software is developed for the data alignment and comparison， with which the product form and position accuracy measurement and the assembly deformation analysis are realized. Finally， the feasibility of the photogrammetry method is verified by an example of a super-large riveted cabin.

Abstract:In order to meet the needs of emergency launch， the storage requirements and performance of key satellite components， e.g.， satellite electronic components， satellite structure， deployment devices， storage battery， and propulsion system， are analyzed in this paper. Taking 6 months’ storage time of the satellite in the launching site as the boundary， the long-term and short-term storage methods for satellites in the launching site are studied respectively. In the short-term storage scheme， the satellite is stored in the whole satellite state， and all components are installed in place. During the storage， the whole satellite is fully powered every 3 months. At the same time， in order to ensure the safety of test and product， the ground test system shields all initiating instructions of explosive devices and opening instructions of push valves， and connects the protection plug of the star meter plug. In the long-term storage scheme， the satellite is stored in the state of platform plus parts， and there are separate requirements for the storage conditions and inspection conditions of each part. Finally， the rapid responses of the storage satellite are analyzed. Up to now， there have been nearly ten successful short-term storage cases for satellites in the launching site and long-term storage cases up to seven years for satellites on the ground.