Research On Modeling, Navigation And Control Technologies For Loitering Munition

Launched by various weapons platform, loitering munitions is a new type of ammunition cruising over the target zone, to perform varieties of combat missions such as reconnaissance and target designation, precision strike, damage assessment, communication relay, airborne early warning, etc., and has already become a hotspot in the development of military facilities. To meet the design requirements of loitering munitions with low cost, miniaturization and long-time hover, as well as to achieve the purpose of autonomous flight, the main work of this paper is to make a deeply research of certain key technologies including loitering munitions system modeling, low-cost integrated navigation, aerodynamic parameter identification, design and analysis of flight control system.Firstly, the six degrees of freedom motion equation of loitering munitions is established. Through decoupling and linearization, the linear model of the longitudinal and lateral motion can be obtained, and the motion mode of loitering munitions is analyzed. The simulation platform of loitering munitions with parameter uncertainty is established. Through acting square wave to the elevator, the proposed linearized equation of motion is verified, besides, the influences of parameter uncertainty on the system are analyzed at the same time.In order to meet the range requirement of various gyroscopes, which operate in different stages of loitering munitions, and to reduce the random drift and noise of MEMS gyroscope, a fuzzy-logic based compensation algorithm for gyroscope array with variable ranges is proposed. Numerical simulation results show that the fusion algorithm can make a choice of gyroscope ranges and can adjust the approach of fusion for the gyroscope array output in real-time, which improves the output precision. To support long-term and high-fidelity navigation data, GPS/MIMU integrated navigation algorithm based on strong tracking Kalman filter with time-varying measurement noise is presented. To reduce the impact on the calculation of the integrated navigation caused by the change of the GPS output error, the measurement noise in this algorithm changes with the DOP values in real-time, which can achieve on-line estimation and compensate the error of MEMS inertial sensors. The vehicle test verifies the effectiveness of the algorithm. Furthermore, when the GPS is disabled, an attitude filtering algorithm of time-varying gain based on gradient descent method is proposed, which can solve the problem that the attitude calculation results could be divergence gradually due to the output error of low cost gyroscopes. Regarding the output of accelerometer and magnetometer as observed values, the attitude error compensation can be achieved by time-varying gain. Simulation and flight test results verify that the algorithm proposed in this paper outperformed Kalman filter algorithm in terms of the analysis accuracy of attitude in the dynamic environment.Furthermore, the longitudinal aerodynamic parameter identification model of loitering munitions is established, and the longitudinal aerodynamic parameters are estimated by various domain methods. The results show that UKF can obtain aerodynamic parameters precisely in an on-line way. In order to solve the problem of lacking the angle of attack and the angle of sideslip measurements in the aerodynamic parameter identification, a three-dimensional wind speed estimation method based on Kalman filter is proposed, which uses the triangular relationship of wind speed, ground speed and space velocity. Once the online estimation of the three-dimensional wind is accomplished, the estimate values of angle of attack and sideslip can be achieved. To ensure safety, a two-stage closed-loop on-line identification strategy is presented, which combines Recursive Least Square(RLS) and UKF. Adding continuous control signal "3211" at the end of the input, and regarding the identification results from RLS stage as the initial value of UKF for the high precision identification of aerodynamic parameters, then the aerodynamic parameters can be identified in the closed-loop condition in a direct way, which avoids the modeling problem for complicated control systems and achieves the initial value for UKF. Semi-physical simulation results show that the strategy can realize the aerodynamic parameter identification in the closed-loop condition.Finally, aiming at the uncertainty of external disturbance and parameter changing in loiter munitions, the control loop is constructed with the weighted function selected and the robust controller is designed using the control theory. The miniaturization and low-cost hardware system is developed for verifying the navigation and control system and has been certificated by semi-physical simulation experiments, which illustrates that the controller has strong robustness to the wind disturbance and parameter uncertainty.