A solid strap-on launch vehicle joint swing control scheme is proposed, which is dedicated to handle large range thrust variation and the non-synchronous thrust among solid boosters. Firstly, the combustion chamber pressure is used to estimate the real-time thrust of each solid booster, and the controller parameters are accordingly modified on-line due to the thrust variation. Secondly, by adopting the feed-forward control to eliminate the loss of thrust synchronism. Finally, the real time swing subsidence angle compensation for solid booster is adopted to solve the effect of pivot point excursion. The simulation results show that the proposed adaptive control scheme is able to efficiently solve control matters caused by the thrust uncertainties and the non-synchronization thrust among solid boosters, which takes advantage of increasing the attitude control precision with large scale and improving the launch vehicle’s adaptive capability in the real flying environment.
In order to solve the problem that the filtering result of RIMU/GNSS integrated navigation system based on filtering method is unstable under the condition of large maneuvering of carrier, an estimation method based on factor graph is proposed. The state equation and measurement equation oriented to the redundant inertial navigation system are established, and the redundant inertial data are fused into the carrier coordinate system three axes. A data fusion method based on factor graph is established, and the fused inertial data and satellite information are abstracted as factor nodes, and the state information is abstracted as variable nodes. The cost function including inertia factor and GNSS factor is established and the state quantity is estimated in a nonlinear optimization way. The reslult of numerical simulation shows that performance of using the factor graph method can effectively reduce the position error of the carrier and the root mean square error is obviously degraded than that of the Kalman filter.
A study is conducted on the problem of achieving attitude safety control under low redundancy in micro-nano satellite. A resource efficient attitude safety control scheme is proposed, which aims to cut off all potential fault sources as much as possible when diagnostic information is not in complete. Only magnetometer and magnetic torquer are used to reconstruct the satellite attitude control system. The proposed scheme can realize angular velocity damping under any initial attitude condition of the satellite, and can control the satellite maneuvering to the sun oriented attitude, which ensures the efficiency of the solar array charging and thus extends the satellite’s lifespan under fault conditions. Mathematical simulation shows that even when the main control components such as gyroscopes, star sensors and momentum wheels are not available, the proposed scheme can still achieve stable single axis safe attitude control towards the sun, which achieves attitude determination precision by better than 2 degrees and the sun orientation precision by better than 5 degrees that greatly improves the satellite’s safety operation ability in orbit.
An adaptive sliding mode control method is proposed for the control problem of a quadrotor UAV composed of six degrees of freedom under model uncertainties and unknown disturbances. This method can be used to realize position and attitude tracking control of quadrotor UAV. Firstly, based on the dynamics quadrotor system, the fully actuated and under-actuated subsystems are divided. The model uncertainties and unknown disturbances without upper and lower limit constraints are fully considered, and the lumped disturbance terms of each subsystem are extracted; Then, with the help of radial basis function neural network, the real-time approximation and estimation of the equivalent controller containing the extracted lumped disturbance terms are inplemented in progress. At the same time, adaptive control method is used to estimate the approximation error term. The sliding mode controller with the estimated values of the equivalent controller and approximation error terms and the corresponding adaptive update laws are designed. Based on Lyapunov theory, the reachability and convergence of the sliding mode surface where the state trajectories of each subsystem lies are analyzed and introduced. Finally, the comparative simulations are used to verify the effectiveness of the proposed method fairly well.
An online tuning method for a three-loop autopilot based on adaptive control theory is presented in this paper. Due to the rapidly changing disturbance of a dolphin head shaped missile in boost phase, the autopilot is unable to follow commands accurately, which overloads by pitch disturbance torque. In order to solve the problem, firstly, the state equation of the three-loop autopilot is established based on short period dynamics, and the pole placement method is applied to design the control parameters in the nominal state. Then, the nominal autopilot is designed as a reference model, and the control parameters are designed by a Lyapunov function using the model reference adaptive method. Aiming at enhancing the stability of the adaptive controller under modeling uncertainties, the projection operators and dead zones are introduced. Finally, the three-loop adaptive autopilot tracking simulation is conducted for the dolphin head shaped missile in boost phase. The simulation results of the three-loop adaptive autopilot show fairly good tracking performance of the dolphin head shaped missile.
In order to solve the problem of positioning and reliable recovery of some important structures of aircraft after flight test, a highly reliable positioning system based on Beidou and GPS dual-mode module is designed. The system uses the BDM 910 module integrated with GPS/RNSS (Radio navigation satellite system) module and RDSS (Radio determination satellite system) module as the core positioning device is used in the system, and the BDS (Beidou navigation satellite system) frequency point and GPS frequency point navigation messages are received through dual links and four antennas redundancy. While effective satellite navigation signals are captured, the system own position information is resoved in real time. The location information is sent to the search device in real time in the form of Beidou short message, and the landing point coordinates of the aircraft are finally determined. The problem of positioning and recovery of relevant important structures is solved in the flight test in a complex environment. The positioning system has the ability to analyze the positioning data and the communication information of the positioning device in real time, the data analysis rate is less than 5 s, the positioning accuracy is less than 10 m, the communication success rate is more than 95%, the system is continuously running and lasts more than 4 hours, and the positioning is highly reliable.
Aimed at the problem that traditional target tracking methods require prior position information of the target which are not suitable for non-cooperative target tracking, a target tracking method based on wide-angle camera image error for video satellite is proposed. A wide-angle camera imaging model is firstly established, and then the difference is calculated between the projected position of the target in the camera image plane and the expected position. The relative error, angular velocity and attitude error between the target and the small video satellite are calculated by image error, which are used as feedback terms to design a nonlinear controller. Finally, the stability of the system is demonstrated by using the Barbalat’s lemma. The simulation results show that the target driven to the desired position in the image plane can be effectively controlled in a wide field of view of a wide-angle camera and the effectiveness of this method is verified by Tiantuo-2 satellite on orbit data.
In this paper, a novel predefined-time line of sight (LOS) angle rate convergence guidance law is proposed for the diverse engagements in three-dimensional (3D) scenarios. By using this guidance law, the LOS angle rate can be converged to zero in an arbitrarily prespecified time, which is independent of the initial conditions and control parameters. Besides, there exists an exact upper bound of the convergence time. Firstly, a planar guidance law is designed to introduce guidance commands in two-dimensional (2D) scenarios by decoupling the 3D engagement model. Then, the guidance commands can be functional in 3D scenarios by the theoretical analysis. Finally, simulations are implemented to validate the guidance law and show the effectiveness of proposed method.
In order to solve the path planning of unmanned aerial vehicles (UAVs) in three-dimensional space under threats, an improved snake optimization (ISO) algorithm is proposed for the challenging optimization of UAVs path planning in three-dimensional complex environments. Firstly, the total cost function including range, threat, altitude and smoothing costs is established, and the path planning of UAVs is transformed into an optimization issue that takes into account the safe operation requirements of UAV. In the framework of snake optimization (SO), Tent chaotic map is used to initialize the population, and the temperature threshold is dynamically adjusted to improve the local optimization ability of the algorithm, and Lévy flight strategy is introduced in the exploitation phase to improve the global optimization ability of the algorithm. The simulation results show that the ISO algorithm is superior to the SO algorithm by solving the path planning of UAVs.
In order to achieve rapid and accurate recognition of combat intent in complex battlefield environments with fewer air combat data samples, a combat intent recognition model based on meta-metric learning framework is proposed.Regarding received small sample data during air combat intent recognition, a two-way gated recurrent unit network based on the air combat temporal data are developed to realize effective feature extraction, and the attention mechanism is introduced for promoting the network to fully extract the temporal core features of the air combat data when facing the small sample data are merely available for obtaining the inter-class differences which finally achieve a fairly higher recognition accuracy and recognition speed.Simulation results show that the model proposed in this paper has better accuracy and real-time performance for air combat target intent recognition, especially in the case of small sample data.
Aiming at unexpected the sudden change in inertia parameters of spacecraft on-orbit servicing, real-time identification methods of inertial parameters based on on-orbit attitude and control information are proposed, and recursive least squares (RLS) and extended Kalman filter (EKF) identification algorithms are designed. Adaptive forgetting factors are introduced into the RLS algorithm, which assign weights of prior data and current data during each iteration process to ensure the identification value tracking in time. In the EKF algorithm, the influence of parameter variation on the prior prediction covariance is defined, and it is substituted to update the prediction covariance matrix to solve the mutation of inertia parameters. Under consideration of the mutation in inertia parameters, the simulation results show that the identification accuracy of RLS and EKF can reach 1.5% and 1%, and the identification time is better than 30 s and 40 s respectively; In the scenario of time-varying inertia, both methods can achieve on-orbit identification of inertia parameters, and the identification accuracy fulfills the requirements of the attitude control system.
In order to fulfill the demands for rapid and precise attitude adjustment during the docking process of solar arrays, and to resolve the challenges associated with manual adjustment,this study proposes a leveling strategy, which can update the kinematic parameters of the AAM according to the ground leveling situation, uses the laser tracker to obtain the pose required for leveling, applies closed-loop joint position control based on kinematic inverse solution to realizing the leveling of the moving platform, avoids the false leg in the leveling process through implementation of the traditional outrigger lifting and leveling method, and realizes the automatic adjustment of the AAM from the state of the outrigger stowed away to the state of touching the ground as support and then to the state that the moving platform is automatically adjusted to the horizontal plane level. The feasibility of the leveling strategy is verified through simulation, which provides a solution to false leg-leveling in the engineering application of AAM during solar wing docking.
