Predictive vehicle speed trajectory optimization

Ye, Ziqi; Pischinger, Stefan (Thesis advisor); Abel, Dirk (Thesis advisor)

Aachen (2019)
Dissertation / PhD Thesis


Climate-related regulations and “new players” engaging the automotive industry such as giants out of consumer electronics and software industry, as well as manufacturers from emerging markets are transforming the established automotive industry in the future. Being “electrified”, “autonomous”, and “connected” are among the major trends for the transformation [1]. With this background, this work introduces a predictive function to optimize the vehicle speed trajectory, in order to reduce the energy consumption of an electrified vehicle by using the benefit of connected and autonomous powertrain technology. Dynamic Programming with forward recursion is adopted to solve the optimization problem under constraints under spatial and time domain. The consideration of boundary conditions in multi-domains enables the optimized vehicle speed trajectory to react to sophisticated traffic conditions. In order to demonstrate the real-world benefits, the function is implemented in a compact class battery electric vehicle with rapid prototyping control unit and evaluated with Model in the Loop environment as well as on the test track for checking the real-time performance. With the test cases selected in this work, the function can achieve 17.2% of energy consumption reduction considering the traffic in an inner-city drive, and 3.3% in a highway drive. One of the major challenges is the validation of the function in all environmental conditions, which requires tremendous effort to generate massive amount of reproducible tests.