A Survey of Model Predictive Control Development in Automotive Industries

Abstract

Model Predictive Control (MPC) technology has been progressed steadily over 30 years ago. MPC applications are currently showing a stable architecture in many industries "e.g. Chemicals, Aerospace, Defense, and Food processing". Recently, the work area of MPC's technologies is changing fast, so it's difficult to preserve track of the rapid progress in industrial applications and academic research. MPC is currently showing huge potential for being used in automotive applications as vehicles subsystems will be progressively coordinated to enhance fuel economy and safety. Thus, novel chances for MPC will arise, including coordination of braking action and powertrain in torque vectoring, coordination of engine functionality and transmission to enhance fuel consumption and reaction, control of complex engines. Being used at a supervisory level, more interaction of MPC with the driver is expected. Thus, a major challengeable research experiment for MPC will be to include a driver prediction model. This is already an ongoing effort. In the other side, the active front steering (AFS) braking controllers have been developed with a more detailed prediction model of the driver steering behavior. Accordingly, the recent developed controllers are guaranteeing the same stability performance, are much more predictable to drive. Likewise, an energy controlling approach is recently suggested, where the driver behavior is modelled as a Markov Chain learned in real-time, and used in a stochastic MPC algorithm. The resulting strategy adapts to the way the car is driven, to the drive cycle, and to the environment, achieving economic fuel consumption close to the one obtained with future information.