Optimal Regulation for Dynamic Hybrid Systems Based on Dynamic Programming in the Case of an Intelligent Vehicle Drive Assistant

This contribution proposes an application of an optimisation algorithm to be implemented in an intelligent drive assistant. A vehicle dynamic model is introduced and after the characterisation of its considered nonlinearities some elements of the background concerning "Dynamic Programming" are given together with the proposed algorithm. A two-point boundary value optimization problem in term of velocity is proposed with its solution using the well-known "Dynamic Programming" method. The novelties of this approach consists of a technique which leads to a global optimum. In fact, this technique searches throughout the whole state-space for it. Because of the nonlinearity of the considered problem, using numerical methods seems to be the only way solving it. Using "Dynamic programming" such kind of a complex problem is solved by dividing it into a collection of simpler subproblems. At the end, an optimal feedback controller for the angular position of the throttle valve of the engine is proposed. Simulation results are discussed together with possible application aspects such as "Curse of Dimensionality" and an explicit analysis concerning the calculation effort.