Reconfigurable Integrated Vehicle Stability Control Using Optimal Control Techniques

Abstract

The motivation for the development of vehicle stability control systems comes from the fact that vehicle dynamic behavior in unfavorable driving conditions such as low road-tire adhesion and high speed differs greatly from its nominal behavior. Due to this unexpected behavior, a driver may not be successful in controlling the vehicle in challenging driving situations based only on her/his everyday driving experience. Several noteworthy research works have been conducted on stability control systems over the last two decades to prevent car accidents due to human error. Most of the resultant stability controllers contain individual modules, where each perform a particular task such as yaw tracking, sideslip control, or wheel slip control. These design requirements may contradict each other in some driving scenarios. In such situations, inconsistent control actions can be generated with individual modules. The development of a stability controller that can satisfy diverse and often contradictory requirements is a great challenge. In general, transferring a control structure from one vehicle to another with a different drivetrain layout and actuation system configuration requires remarkable rectifications and repetition of tuning processes from the beginning to achieve a similar performance. This can be considered to be a serious drawback for car manufacturing companies since it results in extra effort, time, and expenses in redesigning and retuning the controller. In this thesis, an integrated controller with a modular structure has been designed to concurrently provide control of the vehicle chassis (yaw rate and sideslip control) and wheel stability (wheel slip ratio control). The proposed control structure incorporates longitudinal and lateral vehicle dynamics to decide on a unified control action. This control action is an outcome of solving an optimization problem that considers all the control objectives in a single cost function, so integrated wheel and vehicle stability is guaranteed. Moreover, according to the particular modular design of the proposed control structure, it can be easily reconfigured to work with different drivetrain layouts such as all-wheel-drive, front-wheel-drive, and rear-wheel-drive, as well as various actuators such as torque vectoring, differential braking, and active steering systems. The high-level control module provides a Center of Gravity (CG) based error analysis and determines the required longitudinal forces and yaw moment adjustments. The low-level control module utilizes this information to allocate control actions optimally at each vehicle corner (wheel) through a single or multi-actuator regime. In order to consider the effect of the actuator dynamics, a mathematical description of the auction system is included in distribution objective function. Therefore, a legitimate control performance is promised in situations requiring shifting from one configuration to another with minimal modifications. The performance of the proposed modular control structure is examined in simulations with a high-fidelity model of an electric GM Equinox vehicle. The high-fidelity model has been developed and provided by GM and the use of the model is to reduce the number of labor-intensive vehicle test and is to test extreme and dangerous driving conditions. Several driving scenarios with severe steering and throttle commands, then, are designed to evaluate the capability of the proposed control structure in integrated longitudinal and lateral vehicle stabilization on slippery road condition. Experimental tests also have been performed with two different electric vehicles for real-time implementation as well as validation purposes. The observations verified the performance qualifications of the proposed control structure to preserve integrated wheel and vehicle chassis stability in all track tests.