結合觀測器與微分器技術於機動目標加速度估測

Abstract

目標機動是影響導引準確度的關鍵因素之一，本論文利用擴張狀態觀測器與Super-Twisting演算法設計觀測器等方式，實現穩定而精確的目標跟蹤以供後續設計導引律時使用。我們針對實際導引系統中尋標器迴路之量測雜訊干擾，設計濾波器減弱量測雜訊。此外，本論文基於擴張狀態觀測器與Super-Twisting演算法等方式來設計微分器，實現相位超前用以彌補濾波所導致之量測輸出的相位損失。模擬結果顯示所提之結合濾波器與微分器之補償機制能有效降低輸出量測雜訊對估計結果的影響，說明了此結合技術之有效性。另一方面，本論文基於Super-Twisting演算法，利用其收斂時間估計來選取最佳化參數，所獲得的結果提供觀測器可以快速跟蹤至目標訊號。

Target maneuvering is one of the key factors that affects guidance accuracy. This paper applies an extended state observer and a Super-Twisting algorithm observer (STO) to achieve a stable and precise target tracking for the follow-up guidance law design. For the situation that the measurement noise of seeker in the practical guidance system is large, a linear filter is designed to attenuate the measurement noise. Besides, a differentiator which is composed of extended state observer and sliding mode technique is also presented to achieve the phase advance to cover the phase lag caused by the filter. Moreover, a methodology to select a set of better parameters for STO is presented. Simulation results demonstrate the benefits of the proposed scheme.