Improved Gauss Pseudospectral Method for Solving Nonlinear Optimal Control Problem with Complex Constraints

孙勇,张卯瑞,梁晓玲
DOI: https://doi.org/10.3724/sp.j.1004.2013.00672
2014-01-01
ACTA AUTOMATICA SINICA
Abstract:An improved Gauss pseudospectral method(IGPM) is proposed to solve the nonlinear optimal control problem with complex constraints,which is difficult to get the analytical solution.Traditional methods can hardly solve these problems by getting some parameter values from the look-up table.Based on the traditional Gauss pseudospectral method,the IGPM is put forward by transforming the nonlinear final state integral constraints into linear type.The costate obtained by the costate mapping theorem is used to check the optimality,which makes the IGPM have the same accuracy with the indirect method.Besides,the computation methods of the initial costate and the boundaries controls are proposed.In order to increase the accuracy,an iterative algorithm is given based on the IGPM.Finally,the ascent trajectory optimization problem of the hypersonic vehicle is solved by the proposed algorithm.The results show that the optimal trajectory almost satisfies the path constraints and the optimality.
What problem does this paper attempt to address?