Reinforcement-Learning-Based Dual-Control Methodology for Complex Nonlinear Discrete-Time Systems With Application to Spark Engine EGR Operation | IEEE Journals & Magazine | IEEE Xplore