Design and Control Optimization of Hybrid Electric Vehicles: from Two-Wheel-Drive to All-Wheel-Drive Vehicles

Abstract

Fuel efficiency standards in the ground transportation sector have been becoming more stringent over the previous decade worldwide. The power-split hybrid powertrain technology is one of the most promising solutions to meet those exigent standards. This technology has been successfully implemented in the passenger vehicle market, such as Toyota Prius, and demonstrated a fuel economy improvement of over 60%. In contrast, however, few hybrid electric light trucks are available, which is problematic given the fact that trucks are now more than 65% of light-duty vehicle sales in the United States. Additional performance requirements such as all-wheel-drive (AWD) also has not been explored adequately. Expanding the power-split hybrid technology to a broader market while satisfying all these requirements is imperative but challenging. The main contributions of this dissertation includes, 1) we present a systematic design methodology that enables the exhaustive search of AWD power-split hybrid powertrains; 2) the concept of relaxed optimization for additional fuel reduction; 3) a systematic framework of control design that enables automated development of real-time control strategies and ensures near-optimal performance; in addition, 4) an experimental study to verify the theoretical development. Designing AWD power-split hybrid powertrains involves searching over a large design space. Millions of designs are possible when considering collocations of all components including planetary-gear (PG) sets, an engine, electric motor(s), and clutches. Within the developed systematic design methodology, all possible designs can be generated through an automated modeling technique; exhaustively searching through all these designs then become possible. A systematic screening process is developed to screen for feasible designs, with respect to desired performance attributes; optimal designs then can be identified by checking their launching/towing performances together with fuel efficiencies. A case study on an imagined hybrid F-150 light truck demonstrates that the developed methodology is able to identify dozens of better designs than parallel-hybrid baseline model. Optimization is crucial for both design and control development. An optimization of hybrid electric powertrain is defined which allows load leveling among the power source (engine), electrical energy buffer (battery). Relaxed optimization is further defined and investigated when the mechanical energy buffer (vehicle kinetic energy) is also introduced. Analysis of these optimized results are used for design screening and control development. By understanding the analysis of optimized results, a systematic framework is developed to generate a near-optimal real-time control strategy. A set of optimal controls is generated by analyzing the hybrid powertrain system firstly; the real-time control strategy is developed by constructing the policy from the optimal control set. Near-optimal results are achieved under this development framework. With the establishment of the design and control development frameworks, an experimental study is performed to verify this theoretical development. Preliminary results project that the developed framework of hybrid technology implementation is able to identify designs achieving fuel consumption reduction of more than 50% compared to current conventional baseline models for truck applications.