Parallel Hybrid Electric Vehicle Powertrain Modeling and Control for John Deere - 9370 R Tractor

The objective of this paper is to study the suitability of the parallel hybrid electric vehicle (HEV) type power-train for a real tractor - the John Deere 9370 R tractor. The HEV power-train model is developed in MATLAB and Simulink environment using its add-ons such as control systems and Sims-cape libraries. The HEV power-train model contains two power sources: 1. the engine is modeled using power vs. speed and torque vs. speed look-up tables, and 2. the electrical system is modeled using Simulink library. A single motor is utilized for the both purposes, to start the engine and to supply additional power to the wheels when required. This motor also works as a generator and charges the Li – ion battery in two modes known as cruise mode and parking/ charging mode. The power split device as we see in the modern HEVs such as Toyota Prius is replaced by a clutch and gear mechanism to add the torques through a pair of parallel shafts, like trans-axle of Ford Escape HEV model. The lock-up clutch of the torque converter and the motor clutch engages according to the power requirement at wheels and the battery state of charge (SOC%). The gear mechanism is synchronous cone and hub type, as we see in the manual gear box, is used to switch power flows in three ways like power sources to wheels, engine to the generator, and motor to the engine to start the engine. In the case of off-road applications, we require different torque outputs at the draw bar and wheels of the tractor. Hence, a 5 – speed automatic transmission is used in between the sources and final drive. Furthermore, the model is controlled using three PI – controllers, a vehicle speed PI – controller, which is a central controller of the system and two sub PI – controllers to control engine speed and motor torque. The model is tested in 3 three simulations. Firstly, the response of the system is studied to a step input signal. Secondly, the effect of the road grade angle on the performance of the system is investigated. Finally, the efficiency of PI controllers is observed by studying the behavior of the system using a drive cycle.