A Nonlinear Model Predictive Controller With Multiagent Online Optimizer for Automotive Cold-Start Hydrocarbon Emission Reduction

In this paper, a nonlinear model predictive controller (NMPC) with a multiagent heuristic optimizer, which is called dynamic particle swarm optimization (DPSO), is proposed to reduce the cold-start hydrocarbon (HC) emission of an automotive spark-ignited engine. In general, the cold-start HC emission reduction has been proven to be a very challenging control problem that has attracted increasing attention from the automotive research community. The main contribution of this paper lies in the implementation of a model-based predictive control scheme that uses a discrete nonlinear control-oriented model for calculating the control commands. In the first part of the experiments, the developed control-oriented model is validated with some data obtained through experiments. Then, by applying the controller to different cold-starting conditions, the values of exhaust gas temperature Texh and engine-out HC emissions are controlled to reduce the cumulative tailpipe emissions HCcum. To ascertain the veracity of the proposed control scheme, the same problem is solved using Pontryagin's minimum principle. The conducted simulations indicate the applicability of DPSO as an effective solver at the heart of the NMPC. Moreover, the results of modelin-the-loop simulation and hardware-in-the-loop testing demonstrate the acceptable performance of the proposed control scheme for real-time applications, which is based on the NMPC method for the considered problem.