Solving the Parallel Parking Problem in Minimum Time Using a New Optimization Package: COPILOTS

• Autor(a) principal: Lobato, Fran Sérgio
• Data de Publicação: 2021
• Outros Autores: Iasbeck, Arthur Henrique, Assis, Pedro Augusto Queiroz de
• Tipo de documento: Artigo
• Idioma: por
• Título da fonte: Vetor (Online)
• Texto Completo: https://periodicos.furg.br/vetor/article/view/12869
• topic: COPILOTS; Parallel Parking; Optimal Control Problem; Direct Collocation; Optimization; COPILOTS; Estacionamento Paralelo; Problema de Controle Ótimo; Colocação Direta; Otimização

This work aims to present (BasiC OPtImaL COnTrol Solver), a proposed package for solving Optimal Control Problems (OCPs) based on Direct Methods, easy to use and ideal for users with little or no experience in solving PCOs. In general, the proposed tool performs the discretization of controls via trapezoidal colocation or Hermite-Simpson colocation, while cubic polynomials are used in the interpolation of the states. In effect, the original problem (algebraic-differential) is transformed into a purely algebraic problem, which is solved considering the SQP (Sequential Quadratic Programming). For validation purposes, was used to determine the minimum time path to be traveled by a vehicle when executing a parallel parking maneuver and a new algorithm for initializing states and controls was proposed in order to minimize the computational effort spent. The results demonstrate that the proposed package is configured as an alternative for the resolution of OCPs.