IEEE TRANSACTIONS ON POWER ELECTRONICS, VOL. 21, NO. 3, MAY 2006 779 Sliding Mode Based Powertrain Control for Efficiency Improvement in Series Hybrid-Electric Vehicles Metin Gokasan, Member, IEEE, Seta Bogosyan, Senior Member, IEEE, and Douglas J. Goering Abstract—This study involves the improvement of overall effi- ciency in series hybrid-electric vehicles (SHEVs) by restricting the operation of the engine to the optimal efficiency region, using a con- trol strategy based on two chattering-free sliding mode controllers (SMCs). One of the designed SMCs performs engine speed con- trol, while the other controls the engine/generator torque, together achieving the engine operation in the optimal efficiency region of the torque-speed curve. The control strategy is designed for appli- cation on a SHEV converted from a standard high mobility mul- tipurpose wheeled vehicle (HMMWV) and simulated by using the Matlab-based PNGV Systems Analysis Toolkit (PSAT). The per- formance of the control strategy is compared with that of the orig- inal PSAT model, which utilizes PI controllers, a feedforward term for the engine torque, and comprehensive maps for the engine, gen- erator and power converter (static only), which constitute the aux- iliary power unit (APU). In this study, in spite of the simple mod- eling approach taken to model the APU and the optimal efficiency region, an improved performance has been achieved with the new SMC based strategy in terms of overall efficiency, engine efficiency, fuel economy, and emissions. The control strategy developed in this work is the first known application of SMC to SHEVs, and offers a simple, effective and modular approach to problems related to SHEVs. Index Terms—Auxiliary power unit (APU), engine optimal ef- ficiency operation, series hybrid electric vehicles (HEVs), sliding mode controller (SMC). I. INTRODUCTION H YBRID electric vehicles (HEVs) offer the most economi- cally viable choices in today’s automotive industry, while also providing solutions for high fuel economy and very low emissions. The most common HEVs take on two major forms; series and parallel HEVs. The drawbacks of the series HEVs (SHEVs) over parallel types are the requirement for two elec- trical machines and larger dimensions for the traction motor, in addition to the losses incurred during the conversion of the mechanical to electrical energy and back to mechanical energy again. On the other hand, the mechanically decoupled structure Manuscript received March 31, 2005; revised October 26, 2005. This work was supported by the U.S. Army’s Tank-Automotive and Armaments Command (TACOM), its Automotive Research Center (ARC) Program, the Yuma Proving Grounds, the Army’s Cold Regions Test Center (CRTC), and the NSF-CISE Program. Recommended by Associate Editor J. Shen. M. Gokasan is with the Electrical-Electronics Engineering Facility, Istanbul Technical University, Istanbul, Turkey. He is also with the Electrical and Computer Engineering Department, University of Alaska, Fairbanks, AL 99775 USA. S. Bogosyan is with the Electrical and Computer Engineering Department, University of Alaska, Fairbanks, AK 99775 USA (e-mail: s.bogosyan@uaf. edu). D. J. Goering is with the Mechanical Engineering Department University of Alaska, Fairbanks, AK 99775 USA. Digital Object Identifier 10.1109/TPEL.2006.872373 of the engine from the wheels in SHEVs also presents some ad- vantages; i.e., the use of electric motors alone for traction gives rise to the possibility of very low operating noise, offering ben- efits particularly in military operations; additionally, higher ef- ficiency levels can be achieved for the engine by not allowing its operation outside of its optimal efficiency region, except for short transient durations. For more information on HEVs and electrical vehicles (EVs), the readers is referred to [1]–[5] ad- dressing many important aspects of the technology involved and [6] and [7] for the modeling and control of power trains in both parallel and series type HEVs. The SHEV powertrain considered in this work consists of a battery bank and an engine-generator set which is referred to as the auxiliary power unit (APU), two traction motors, and power electronic circuits to drive the generator and traction motors. The control unit of the powertrain executes two main control algorithms. 1) Control of the traction motors, which involves the delivery of the torque value (demanded by the driver) to the wheels. 2) Control of the APU, which is particularly important in terms of efficiency and emissions in SHEVs. The general strategy is based on the operation of the engine in its optimal efficiency region with the consideration of the battery state-of-charge (SOC). By controlling the engine speed and generator torque, the engine is forced to operate in its highest possible efficiency region. In the existing literature related to SHEVs, studies are mostly concentrated on the control of the APU. Among these studies, [8] presents a fourth-order linear adaptive dynamic program- ming method for the control of the APU. This method requires minimal a priori knowledge of the plant. In [9], cascade PI con- trollers are utilized for APU control. The signal obtained with the addition of a feedforward term to the cascade PIs used for engine speed control in the outer loop, performs the torque con- trol of the generator. However, the highly nonlinear nature of the system poses problems in the determination of PI controller pa- rameters and feedforward terms. A simulation program called PSIM is given in [10], presenting simulation results for logic decision mechanisms designed for the control strategy of both series and parallel HEVs. [11] introduces a control rule based on complex calculation methods that make use of the engine effi- ciency map, driver power demand, dc source and battery model. In [12], a Matlab/Simulink based model is derived for the sim- ulation of SHEVs and in [13], this model is used for the simu- lation of the system under different driving schedules. In [14], two algorithms are developed for SHEV control to achieve max- imum energy efficiency by determining the generator ON/OFF period and to produce demanded torque when the generator is 0885-8993/$20.00 © 2006 IEEE