2 nd International Conference on Multidisciplinary Research & Practice Page | 59 Volume III Issue I IJRSI ISSN 2321-2705 Model-Based Design of a Parallel HEV Ankur Baruah, Tanuja Sheorey, Vijay Kumar Gupta Mechatronics, IIITDM Jabalpur Madhya Pradesh, India Abstract—- In this paper a quasi-static model of a parallel HEV powertrain is developed and applied for model based control. An effective set of features have been derived from a city drive cycle to characterize traffic congestion levels. This characterization model was later used with a power management strategy to control the torque split between engine and motor. Simulationswere carried on Indian Urban Cycle and the feasibility andeffectiveness of the control strategy is tested with the stateof charge (SOC) and fuel efficiency. The benefits of the model are demonstrated through simulation results. Keywords - Parallel hybrid, model-based design, PCA-fuzzy control, QSS toolbox I. INTRODUCTION ollution caused by transport is a heavy burden. According to an UN study, India, China and South-East Asia are among the few geographic locations, which are undergoing rapid urbanization, leading to an increase in emission of CO 2 and other Greenhouse gases [1]. Moreover, stringent governmental regulations [Euro 5, Bharat Stage V] around the world have increased the demand of cleaner vehicles. Introduction of hybrid vehicles would help automotive manufacturers meet these regulations. In comparison to electric and fuel cell vehicles, Hybrid Electric Vehicles (HEVs) have proved to be economically viable due to the use of smaller battery pack and their similarities to conventional vehicles [2]. This paper demonstrates a model-based design of a parallel HEV in Matlab/Simulink ® . Mahapatra S. et al. have shown how the model-based approach is viable in overcoming multi- domain complications that arises from the complex interaction between various mechanical and electrical components [3]. Gadda C. and Simpson A. have published an eloquent work by Tesla Inc. on model-based design of the Tesla Roadster [4]. This work emphasizes on a quasi-static approach to model the drivetrain of the vehicle. The quasi-static model simulations are carried out using QSS toolbox that has an integrated environment facilitating the combination of various subsystems [5]. A backward-facing HEV has been implemented, based on the power flow. Fig 1 gives a brief insight into parallel HEV design and model implementation. The work started with analysis of a sample urban drive cycle to find parameters that have a comparatively high impact on improving the overall efficiency of a vehicle. Principal Component Analysis (PCA) has been used to find patterns in the data.The parameters are then compressed without much loss of information [6].The work by Guzzella L. and Sciarretta A. has provided detailed insight into the mathematical modeling and subsequent optimization of vehicle propulsion systems and has been extensively referred in this work [7]. The battery model takes into account the regenerative braking capabilities of an HEV. The design of control algorithm is the key to utilizing the full potential of a hybrid powertrain. Power or energy management acts as a supervisory control algorithm to control the torquesplit between engine and motor. Karbaschian M. A. et al. have put together a review that explains and compares various power management approaches for HEVs [8]. A Fuzzy Logic Controller (FLC) based power management strategy that decides the torque split ratio between the engine and the motor has been implemented in this paper [9]. Fig 1Power flow in a backward-facing Parallel HEV Model In Section II a brief analysis of the drive cycle is provided. Section III focuses on the mathematical framework and modeling of the drivetrain where the modeling of each block is followed by the results pertaining to that particular block. The control strategy is explained in detail in Section IV. Results and conclusion are detailed in Section V. II. DRIVE CYCLE ANALYSIS A. Sample Collection and Representative Features To exploit regenerative braking in hybrid vehicle, a city drive cycle is an ideal choice. The frequent slowing-down, stop and go situations indicate greater chances of regeneration and hence a better fuel economy. Here, Indian Urban Cycle [10] has been used to carry out the analysis and simulation. Sample of Indian Urban Cycle is shown in Fig 2. A review on driving style recognition algorithms provides a summary of parameter used to recognize drive cycles. The study shows that parameters based on velocity and acceleration are the most frequently used [11]. However, too many parameters cause higher hardware cost, longer computational time and the parameters need to balance with the “curse of P