An effective dynamic optimization method based on modified orthogonal collocation and reduced SQP

An effective dynamic optimization solution method based on the modified orthogonal collocation (mOC) and reduced successive quadratic programming (rSQP) is proposed, where the mOC is proposed to decrease the approximation error of the discrete optimal problem while traditional OC method converts the dynamic optimization problem to a regular but discrete nonlinear programming (NLP) problem, and the rSQP method is introduced to solve the resulting NLP problem in the reduced space of the independent variables. A classic benchmark of dynamic optimization problem is explored as demonstration, where the detailed comparative researches between the literature reports and the proposed method are carried out. The research results illustrate the efficiency of the proposed method.