Model predictive control for aftertreatment regeneration

A method for controlling an aftertreatment regeneration process for an aftertreatment system including an aftertreatment system and a hydrocarbon doser operatively coupled to the aftertreatment system, the method comprising:

• receiving an aftertreatment regeneration request;
• receiving temperature information generated by at least one temperature sensor operative to monitor a temperature of the aftertreatment system;
• generating a plurality of estimated internal temperatures of the aftertreatment system for a respective plurality of positions within the aftertreatment system;
• generating a plurality of predicted future states of the aftertreatment system based on the aftertreatment regeneration request, the temperature information generated by the at least one temperature sensor, and the plurality of estimated internal temperatures of the aftertreatment system, wherein the plurality of predicted future states of the aftertreatment system includes at least three predicted future states generated for at least three future time steps;
• and
• generating, based on the aftertreatment regeneration request, the temperature information generated by the at least one temperature sensor, and the plurality of predicted future states of the aftertreatment system, a control command for actuating the hydrocarbon doser.

A model predictive controller for controlling an aftertreatment regeneration process for a machine including an aftertreatment system and a hydrocarbon doser operatively coupled to the aftertreatment system, operative to:

• receive an aftertreatment regeneration request;
• receive temperature information generated by at least one temperature sensor operative to monitor a temperature of the aftertreatment system;
• generate, based on the aftertreatment regeneration request and the temperature information generated by the at least one temperature sensor, a predicted future state of the aftertreatment system;
• and
• generate, based on the aftertreatment regeneration request, the temperature information generated by the at least one temperature sensor, and the predicted future state of the aftertreatment system, a control command for actuating the hydrocarbon doser,
• wherein the control command is generated at least in part using a cost function, and
• wherein generating the control command using the cost function comprises:
• evaluating multiple sets of candidate control points based on one or more soft constraints used to rank the multiple sets of candidate control points;
• selecting a set of control points from the multiple sets of candidate control points that optimizes the cost function;
• and
• generating the control command for actuating the hydrocarbon doser based on the set of control points selected from the multiple sets of candidate control points.