Gliding Trajectory Programming of Guided Projectile Based on Improved L1 Penalty Successive Convex Programming Algorithm
Qinghai Wang, Chen Qi, Zhongyuan Wang · 2021 21st International Conference on Control, Automation and Systems (ICCAS) · 2021
Taking the general control energy optimal trajectory programming model as the research object, the Radau pseudo-spectral method is used to discretize the continuous variables and then linearly convexify the nonlinear dynamic equation to establish the standard convex optimization model. In order to solve the problems of large amount of redundant computation and slow convergence speed of the traditional L1 penalty successive convex programming algorithm (LPSCP), an improved L1 penalty successive convex programming algorithm (ILPSCP) is proposed in this paper. The algorithm introduces the feasibility judgment process to avoid solving the more complex L1 subproblem PL1when the original problem$P$0is feasible. Taking the gliding trajectory model of the guided projectile in the longitudinal plane as an example, the traditional LPSCP algorithm, the ILPSCP algorithm proposed in this paper and GPOPS2 are used to simulate and compare. The simulation results of ILPSCP algorithm are highly consistent with those of GPOPS2, which proves the effectiveness of the proposed ILPSCP algorithm for solving trajectory programming problems, and the convergence rate of ILPSCP algorithm is improved by 46.36% compared with traditional LPSCP algorithm.