Noniterative Convex Programming-Based Optimal Guidance With Field-of-View and Acceleration Limits

Abstract

In this study, a noniterative, convex programming-based optimal guidance is proposed to address field-of-view and acceleration constraints while achieving terminal impact angles. To circumvent the problems of successive convex programming, such as convergence stability and local optimality, the original guidance problem is reformulated as a quadratic programming problem by employing a limited approximation of look angle. Furthermore, a range-to-go weighted cost function is utilized, which allows the acceleration profile to be shaped based on the gain of the weighting function. A pseudo-spectral method is adopted in the transcription process due to its high solution accuracy. The proposed method rapidly provides optimal guidance solutions without iterations and convergence issues, which makes it suitable for real-time implementation. The performance and utility of the proposed method are demonstrated via extensive simulations.