International Journal of Electrical and Computer Engineering (IJECE) Vol. 12, No. 2, April 2022, pp. 1131~1138 ISSN: 2088-8708, DOI: 10.11591/ijece.v12i2.pp1131-1138  1131 Journal homepage: http://ijece.iaescore.com Statistical modeling and optimal energy distribution of cogeneration units by genetic algorithms Othmane Maakoul 1 , Hamid El Omari 1 , Aziza Abid 2 1 Laboratory of Renewable Energies Environment and Development, Faculty of Science and Technology, Hassan 1 st University, Settat City, Morocco 2 Laboratory LaSTI, National School of Applied Sciences, Sultan Moulay Slimane University, Beni Mellal City, Morocco Article Info ABSTRACT Article history: Received Mar 1, 2021 Revised Oct 29, 2021 Accepted Nov 14, 2021 Our main objective is to evaluate the performance of a new method to optimize the energy management of a production system composed of six cogeneration units using artificial intelligence. The optimization criterion is economic and environmental in order to minimize the total fuel cost, as well as the reduction of polluting gas emissions such as COx, NOx and SOx. First, a statistical model has been developed to determine the power that the cogeneration units can provide. Then, an economic model of operation was developed: fuel consumption and pollutant gas emissions as a function of the power produced. Finally, we studied the energy optimization of the system using genetic algorithms (GA), and contribute to the research on improving the efficiency of the studied power system. The GA has a better optimization performance, it can easily choose satisfactory solutions according to the optimization objectives, and compensate for these defects using its own characteristics. These characteristics make GA have outstanding advantages in iterative optimization. The robustness of the proposed algorithm is validated by testing six cogeneration units, and the obtained simulation results of the proposed system prove the value and effectiveness of GA for efficiency improvement as well as operating cost minimization. Keywords: Cogeneration unit Cost of operation Gas emissions Genetic algorithms Optimization Statistical model This is an open access article under the CC BY-SA license. Corresponding Author: Othmane Maakoul Laboratory of Renewable Energies Environment and Development, Faculty of Science and Technology, Hassan 1 st University Settat City, Morocco Email: othmane.1992.mkl@gmail.com 1. INTRODUCTION The global demand for energy has increased strongly in the last decades due to the industrial revolution and the change of lifestyle. On the one hand, the reserves of fossil resources (oil, natural gas, coal) are limited, and on the other hand, the use of these resources is responsible for the increase of greenhouse gas (GHG) concentrations in the atmosphere, leading to global warming [1]. Among the solutions adopted to ensure an energy transition to non-polluting and reliable energies, we find the cogeneration units that help to approve the demand for electricity and heat during peak periods with a total efficiency that exceeds 80% [2], [3], which will significantly reduce the electricity purchase bill following a decrease in the amount of electricity purchased from the supplier. It is a technology that can work with several types of fuel and is independent of climatic conditions [4], [5], unlike other renewable energies that are intermittent, such as solar and wind power [6]. Optimization of power generation sources is becoming a critical necessity in order to increase the efficiency of the generation system while minimizing losses and emissions of pollutants [7]. An optimized