Research On Sliding Mode Control-Based Precise Guidance Technology

Guidance system is the fundamental requirement and important guarantee of tactical missiles to achieve precise interception. On the background of precise guidance against maneuvering targets, this thesis has proposed several robust guidance laws mainly based on sliding mode control(SMC), combined with disturbance observer, numerical differentiator, generalized model predictive control(GMPC), backstepping and adaptive control, respectively. The main results of this thesis are summarized as follows:Considering the fact the some seekers cannot measure line-of-sight(LOS) angular rate directly, this thesis investigates the problem of guidance law design without LOS angular rate measurement. Three different observer-based second-order sliding mode(SOSM) guidance laws have been proposed to address this problem. Firstly, higher-order slidng mode(HOSM) disturbance observer and differentiator, respectively, are utilized to provide accurate estimations of target maneuver and LOS angular rate. Based on the reconstructed information, a novel SOSM guidance law is then synthesized to drive LOS angular rate to zero in finite time. Due to the contibuous property of the proposed guidance law, the undesired highfrequency chattering of classical SMC is mitigated effectively. Secondly, in order to avoid the complex nature of designing disturbance observer and differentiator separately, a novel finite-time convergent state observer that can precisely estimate the LOS angular rate and the lumped uncertainty simultaneously is designed. Then, a new SOSM guidance law is proposed by using this observer. Finally, based on adaptive Super-Twisting algorithm, another SOSM guidance law is presented to achieve finite-time stabilization of LOS angular rate, in which adaptive technique is applied to estimate and compensate the upper bounds of the gredients of the lumped uncertainties, and SOSM presents rapidity and robustness. This guidance law requires no information on target maneuvers, so it can be easily implmented in practice, where such knowledge is typically unavailable.To improve the effect of the warhead, the problem of terminal angle constraint guidance law design is addressed. Inspired by the concept of zeroing LOS angular rate will lead to a perfect interception with zero miss distance, the relationship between terminal impact angle and terminal LOS angle is presented analytically. Based this relationship, a novel inverse optimal guidance law is proposed by using GMPC technique and SOSM disturbance observer, where the GMPC controller is used to obtain optimal LOS angle tracking performance for non-maneuvering target while the SOSM disturbance observer is used to estimate and compensate for the unknown target maneuver. This guidance law requires no information on the remaining flight time or the so-called time-to-go and the target maneuver, which plays an important role in most existing optimal guidance laws. Next, in order to overcome the drawbacks of existing nonsingular terminal sliding mode(NTSM) guidance laws, a continuous finite-time convergent guidance law that can achieve precise interception with desired intercept angle is designed based on NTSM, disturbance observer and SOSM techniques.To improve the overall guidance performance, this thesis considers guidance law design considering autopilot dynamics. By regarding the autopilot dynamics as a first-order system, a finite-time convergent guidance law is proposed based on backstepping-like approach and HOSM disturbance observer. During the derivation of this guidance law, two integral type control Lyapunov functions are adopted, which can completely address the problem of ‘exposion of terms’ caused by repeated analytic differentiation in classical backstepping controller. Then, in order to fulfill the requirement of terminal impact angle, a fast NTSM(FNTSM) surface is designed, which exhibits the merits of both NTSM and conventional linear SMC surfaces. Combining FNTSM technique with generalized disturbance observer, finite-time control and backstepping approaches, a finite-time convergent guidance law with terminal angle constraint has been proposed to effectively compensate for the second-order autopilot dynamics. In order to address the problem of ‘exposion of terms’, tracking differentiator(TD) is utilized in each backstepping step to estimate the derivatives of the virtual control laws.In order to extend the interceptor’s kill envelop and improve the accuracy of missile systems, this thesis is also converns with robust integrated guidance and control(IGC) design. Considering the aerodynamic coefficient and inertia uncertainties, cross-coupling effects between two channels and target maneuvers, a realistic three-dimensional IGC model is constructed first. For a class of multi-input-multi-output(MIMO) uncertain systems, a multivariable adaptive disturbance observer(MADOB) is designed based on SOSM technique to estimate the lumped uncertainties in finite time and its corresponding multivariable adaptive differentiator(MAD) is also presented. By virtue of the proposed adaptive law, the upper bounds of the lumped uncertainties and their gradients are not required in observer design. Based on the reconstructed information provided by MADOB and MAD, a robust IGC law is then synthesized using block-backstepping approach.