Model predictive control systems and methods for increasing computational efficiency

A method for controlling an actuator system within a motor vehicle, the method comprising:

• utilizing an actuator control system that provides a desired output for at least a first actuator of the actuator system;
• converting within a first conversion module the desired output to a desired first actuator position;
• converting within a second conversion module the desired first actuator position to a target first actuator position;
• utilizing a model predictive control (MPC) module having an MPC solver that determines optimal first actuator positions based on a current position of the first actuator;
• determining within a prediction module future state values of the actuator control system using a mathematical model of the actuator system;
• determining within a cost module a first cost for a first possible set of MPC target actuator duty cycle values;
• identifying within a control module optimal sets of target actuator motor duty cycle values for the MPC module;
• and
• determining if the MPC solver has determined optimal actuator motor duty cycle values for the plurality of actuators;
• and
• applying a Linear Quadratic Regulator (LQR) solution for the target actuator motor duty cycle values if the MPC solver fails to determine optimal actuator motor duty cycles for the plurality of actuators, and wherein the MPC module controls operation of an actuator duty cycle to achieve target actuator positions based on target actuator motor duty cycle values.

An actuator control system for a motor vehicle using model predictive control (MPC), the system comprising:

• an MPC module with an MPC solver having a first logic that determines current system constraints based on a current position of an actuator;
• the MPC solver having a second logic that determines a cost for a set of actuator duty cycles for controlling the position of the actuator;
• the MPC solver having a third logic that determines if the MPC solver has a solution for the current system constraints;
• the MPC solver having a fourth logic for applying a Linear Quadratic Regulator (LQR) solution if the MPC solver does not have a solution for the current system constraints;
• and
• the MPC solver having a fifth logic for constraining the LQR solution to physical and programmed limitations of the actuator system,
• wherein the MPC solver comprises a Kalman filter, and a quadratic programming (QP) problem solver.