2018-01-0860 Published 0 3 Apr 2018 © 2018 SAE International; General Motors LLC.Supervisory Model Predictive Control of a Powertrain with a Continuously Variable Transmission Alberto Bemporad and Daniele Bernardini ODYS Srl Michael Livshiz and Bharath Pattipati General Motors LLC Citation: Bemporad, A., Bernardini, D., Livshiz, M., and Pattipati, B., “Supervisory Model Predictive Control of a Powertrain with a Continuously Variable Transmission,” SAE Technical Paper 2018-01-0860, 2018, doi:10.4271/2018-01-0860. Abstract This paper describes the design of a supervisory multi - variable constrained Model Predictive Control (MPC) system for driver requested axle torque tracking with real-time fuel economy optimization that is scheduled for production by General Motors starting in 2018. The control system has been conceived and co-developed by General Motors and ODYS. The control approach consists of a set of linear MPC controllers scheduled in real-time based on powertrain operating conditions. For each MPC controller, a linear model is obtained by system identification with vehicle and dynamometer data. The supervisory MPC coordinates in real time desired Continuously Variable Transmission (CVT) ratio and desired engine torque to satisfy the system require - ments, based on estimates of axle torque and engine fuel rate, by solving a constrained optimization problem at each sampling step. Each linear MPC controller is equipped with a Kalman filter to reconstruct the system state from available measurements. Compared to more classical controls, the presented MPC approach achieves better coordination of powertrain actuators to satisfy system requirements, while maintaining robustness with respect to measurement noise, ambient conditions, and part-to-part variations. Moreover, the systematic, model-based framework developed for produc - tion enables a potential adaptation of the design to different powertrain architectures. Introduction The growing complexity of powertrain control systems, integration with various types of engines, transmis - sions, and vehicle dynamics control systems has led to the use of torque-based powertrain control. Torque-based control enables flexibility and expandability of the powertrain control system structure. It allows various powertrain actua - tion technologies, such as cylinder deactivation, dual clutch transmission (DCT), and continuously variable transmissions (CVT), to be incorporated, and to enable a simpler control structure than current production controls. In an effort to address the ever-increasing demand for improved fuel economy, many CVT applications have been proposed in the automotive domain. The use of a CVT enables good fuel economy performance because the engine can operate at the optimal fuel consumption point more often. At the same time, vehicles equipped with CVT can have better drivability than automatic transmissions because of continuous transmission ratio changes. Among the many recent patents and publications related to the use of CVT technology in automotive, several are dedi - cated to CVT hardware design. Toyota, Nissan, Honda and Jatco have been working on the improvement of CVT hardware to increase its effectiveness as an actuator for powertrain controls [ 1, 2, 3, 4]. Other publications are devoted to improvements of CVT ratio controls and to performance analysis of control systems with CVT. In [ 5] Bosch describes a CVT control strategy where the slip between pulley and belt is considered as the main control parameter, which leads to a decrease of clamping forces and consequently to better fuel economy. In [ 6] Honda describes a CVT control strategy to improve drivability of a car, making powertrain with CVT behaving the same way as powertrain with automatic trans - missions under wide open throttle scenarios. In [ 7] the authors come up with a control strategy to optimize CVT efficiency by analyzing the correlation between CVT variator losses and the nece