Study On Passive Underwater Target Tracking And Adaptive Guidance Law Design

In this dissertation, based on the state of the art, the exiting problems and the practical requirements of the underwater target tracking and guidance law design, the theories and applications of passive underwater target tracking and adaptive guidance law design are deeply investigated. The main contributions of the dissertation are summarized as follows:1. Filter stability theory(1) Observability of passive underwater target tracking is investigated. By utilizing a pseudo-linear structure of pseudo-measurements obtained by algebraic manipulation of the original nonlinear angle measurements, a simple but effective approach to set up necessary and sufficient conditions for observability is presented. As needlessly solving nonlinear equation and Fisher information matrix, it can be easily applied to general target motions used in practice.(2) Characteristic of Kalman filter in rectangular coordinates and modified polar coordinates is analyzed. With the closed-form expression of covariance matrix, the error of bearing and initiation of covariance matrix are verified to be affectable to the elastic behavior of Kalman filter.2. Adaptive Kalman filter algorithm study(1) Based on the study of the passive target tracking in modified polar coordinates, the nonlinear dynamic model is devised. In connection with the problem of slow convergence and low accuracy, a new algorithm with on-line estimator is presented. The combined scheme of own-ship trajectory and MFC filter, using the realistic scenarios logarithmic spiral and circular arc, produces faster convergence in obtaining estimates of radiating source position than original MFC filter.(2) Aiming at the unstable and biased behaviour of Kalman filter in rectangular coordinates, a rotated covariance extended Kalman filter algorithm and adaptive two-step Kalman filter algorithm are presented. Both of the above Kalman filter algorithms overcome the bad affect caused by linearisation and produce faster convergence, higher precision with the scenarios of line forward cruise and crawl forward cruise than original filter.(3) Considering the linear-in-state properties of the Cartesian state variable definition and the linear-in-measurement properties of the Polar state variable definition, an adaptive Kalman filter algorithm is presented. The hybrid approach employs Cartesian system for state andstate covariance extrapolation and polar system for state and state covariance updating. Simulation results show preliminarily that the improved approach can achieve accurate state estimation. 3. Adaptive guidance law study(1) Motion tracking variable structure control theory is applied to design guidance law for torpedo. This motion tracking variable structure guidance(MTVSG) law can force the torpedo-target line-of-sight angular rate to vary on a given motion for enhancing observability of a passive guidance problem. Simulation studies indicate that under the MTVSG, the observability is apparently enhanced and the guidance accuracy is relatively high.- Further studies show that the MTVSG is robust to noise and parameter disturbance.(2) Based on the fuzzy logic, a bearing with noninterference hybrid guidance law is presented to overcome the drawbacks of the traditional hybrid guidance methods and realize the automotive and intelligent guidance.