Symplectic pseudospectral methods for optimal control. Theory and applications in path planning (Q2240977): Difference between revisions

The book presents -- original, efficient, with good accuracy and convergence -- robust numerical methods for solving different nonlinear optimal control problems with dynamics described by ODEs.The methods combine the results from analytic mechanics with optimal control theory. Namely, optimal control problems are transformed into Hamiltonian systems by the Pontryagin's Maximum Principle or the Variational Principle. The phase flow in Hamiltonian systems is a symplectic transformation. So then, symplectic conservative numerical methods are used to solve optimal control problems efficiently. In the book, a series of symplectic pseudospectral methods (SPMs) for different optimal control problems are developed. The book is divided into three parts. In Part I a short review of the numerical methods for nonlinear control problems such as indirect, direct, hybrid and artificial-based methods are presented. Next, the mathematical formulation of problems studied in the book and the necessary knowledge on the Hamiltonian structure of optimal control problems, symplectic theory and pseudospectral methods are discussed. In Part II symplectic pseudospectral methods for nonlinear optimal control problems such as: unconstrained control problems, problems with inequality constraints, time-delayed problems, predictive control problems are developed. In Part III interesting applications in trajectory planning and control related to: optimal maneuver of the spacecraft, trajectory planning for unmanned ground systems with different configurations, control of three-dimensional overhead cranes, trajectory planning for tractor-trailer systems are presented and discussed in details. Examples in the book are shown according to the following scheme: problem formulation, analytic solution, if available, treated as a benchmark for SMP numerical solution, plots presenting both analytic and numerical solutions, discussion and remarks. Results of numerical simulations verify good properties of the presented methods. In the book, the recent progress in symplectic numerical algorithms using pseudospectal methods for optimal control problems achieved by the authors is reported. The book is well-written. It can be recommended to researchers, engineers and graduate students who are interested in automatic control, optimal control problems and their effective solutions.