Model-Based Design of PHEV Adaptive Control

Abstract

The aim of the proposed PHEV adaptive control is to improve real-world fuel consumption. To achieve this, e-Horizon information is utilized to further improve the energy management strategy. Provided that the end destination is known, road information for the entire route becomes available at the beginning of the journey. The proposed technology consists of three main parts: future speed prediction with Markov chain algorithm, dynamic programming optimization and an adaptive control. The usage of long-term and short-term prediction and optimization is introduced to mitigate the uncertainties of realworld driving. This process is handled by the arbitration of the control, which coordinates short-, long-term and the existing rule-based strategies. The complete solution has been developed applying the V-model, starting from concept to implementation and to test. Testing has been conducted to assess fuel consumption and emission improvements compared to a predefined rule-based control. With customer payback analysis, target customer and driving condition are defined and a potential 4 - 40% running cost-benefit is identified.