Model based predictive engine torque control for improved drivability

Abstract

Elasticity of the various driveline components and backlash originating from gear reduction mechanisms and fasteners may cause torsional vibrations resulting in unintended shunt and shuffle behaviours, when a vehicle is subjected to an acceleration change request. As a result of recent improvements to engine control structures and computational capability developments during the last few decades, the idea of using generated brake torque control has been considered a state of the art research topic among academic researchers and original equipment manufacturers (OEMs). In order to improve transient vehicle response to an acceleration change manoeuvre, a novel engine generated brake torque based model predictive control (MPC) algorithm with an additional anti-shuffle control element has been developed to manipulate the pedal map oriented torque demand signal in an automotive powertrain application. A four mass powertrain model was built and model validation considering longitudinal vehicle dynamics was performed with vehicle level tests using a tip-in followed by a tip out acceleration pedal signal input manoeuvre. Comparison of simulation results and vehicle test data shows that the proposed model is capable of capturing the vehicle acceleration profile revealing unintended error states for the specified input signals. MPC structures based on three mass vehicle models (derived from the four mass model via subtracting tyre dynamics due to high order nonlinearities at the tyre model) was developed in a “MATLAB/Simulink” environment to obtain a smooth and responsive acceleration profile. The MPC controller delivers signal states with the expected performance metrics without error states such as excessive jerks and shuffles. An additional engine to wheel speed difference based proportional controller is employed in order to further reduce powertrain oscillations without compromising from overall system response speed resulting in a comfortable drive.

MPC-Based Coordinated Control of Gear Shifting Process for a Power-split Hybrid Electric Bus with a Clutchless AMT

“…According to the control principle of MPC, the predictive horizon and control horizon are defined as m and p , respectively, thereby confirming that m ≥ p . In addition, the following assumptions are made: ① Outside the (23)…”

“…And the advantages of fast response and high control accuracy of the motor can be fully utilized to compensate torque changes in time. At present, researches on the torque response characteristics of engine and motor are quite sufficient [23,24]. By establishing the transient torque response model of power source, MPC was used to optimize the torque distribution of motor and engine in receding horizon, which can effectively suppress the shift shock.…”

Active vibration control of automobile drivetrain with backlash considering time-varying long control period

“…In this study, the cut-off frequency was determined as 10–30 Hz with a sufficient margin from the first natural frequency of 4.0 Hz. The gain in the passband of the low-pass filter becomes a weighting constant in minimizing the H 2 norm 25–27 of z 1 when designing the controller in Figure 4(a). Therefore, designing its value to be too small may lead to an insufficient transient vibration performance.…”

“…8 In particular, many outcomes are related to model predictive control (MPC). 9–14 These are broadly considered to be effective techniques. A mode-switching-based active control algorithm including the sliding mode technique and PID control has been investigated by simulation studies.…”

Double‐Target Switching Control of Vehicle Longitudinal Low‐Frequency Vibration Based on Fuzzy Logic

“…In order to design an advanced model-based controller, a simplified model is necessary to be able to capture the system dynamics. Several driveline and vehicle longitudinal dynamic models have been proposed in the literature, in which two mass models are the most common ones considering benefits of simplicity for running controller algorithms [15][16][17]. is study is based on a two mass vehicle model with the road load component for simulating longitudinal dynamics shown as Figure 2.…”