Despite the potential benefits of the Homogenous Charge Compression Ignition (HCCI) engines, controlling the combustion phasing and emissions are the main obstacles in its development but, in conventional engines the load demand is simply controlled by accelerator. The aim of present work is developing a HCCI engine controller which controls its main parameter similar to conventional engine: controlling the combustion phasing based on load demand. In the present study, A Multi-Input Multi Output (MIMO) controller has been designed and validated which is used to control HCCI engine main parameters such as: optimum combustion phasing, engine load and minimum emissions. Similar to conventional IC engine the controller is capable of tracking all desired set-points solely by means of load demand as the only reference trajectory. A physic-based control model together with application of Artificial Neural Networks (ANN) is developed to predict the optimum combustion phasing with minimum emissions based on load demand. The results of proposed model are validated with the experimental data for steady state and transient cases. The optimal CA50 is selected based on minimizing the emissions using a multi-zone kinetic coupled with a genetic algorithm. The developed controller performance has been tested thoroughly to evaluate the tracking and disturbance rejection capabilities. Results show that it is capable of rejecting the disturbances for fix engine loads. The controller is quick to reject the disturbances within 3-5 engine cycles, while deviations within 0.04 bar, 0.5CAD and 0.03 for IMEP, CA50 and emissions respectively. (C) 2019 Elsevier Ltd. All rights reserved.