Dynamic Analysis and Multivariable Transient Control of the Power-Split Hybrid Powertrain

As an effective approach for improving automobile fuel economy, powertrain hybridization has inspired extensive research efforts on system control and energy optimization. To convert the output of the energy management strategy into actual motion of the engine and the hybrid actuators (motor and generator), the hybrid powertrain controller decides to what extent the optimal operational trajectories of the engine and the hybrid actuators can be realized, so as to achieve the expected fuel efficiency and emissions benefit. To develop a hybrid powertrain controller with higher bandwidth and accuracy, this paper presents the design and experimental validation of a multivariable control framework for the power-split hybrid powertrain, as an improvement over the existing decoupled single-input, single-output (SISO) controller. First, the dynamic analysis characterizes the transient behavior (inverse transient dynamics) of the hybrid powertrain based on its unique physical structure, and reveals the inherent limitations of the SISO hybrid controls on transient engine operation (speed/torque tracking) due to the dynamic interactions among the multiple system variables and the electrical system constraints. Then, to solve this problem and further improve the transient engine operation tracking performance, a multivariable hybrid powertrain controller is designed to dynamically coordinate the engine torque and hybrid torques during transients, so as to flexibly manipulate the transient driving torque of the hybrid vehicle. To validate the proposed control design, a rapid-prototyping hybrid research platform with three-level control architecture is used. Experimental results show the multivariable hybrid powertrain control can produce a faster and balanced engine transient operation, with the moderate electric torque/power usage.