A large number of biological information has been available from genome sequencing and bioinformatics. To further understand the qualities of the biological networks (such as metabolic network) in the complex biological system, representations of integrated function <i>in silico</i> have been widely investigated, and various modeling approaches have been designed, most of which are based on detailed kinetic information except flux balance analysis (FBA). FBA, just based on stoichimetrical information of reactions, is a suitable method for the study of metabolic pathways, and it analyzes the behaviors of the network from the viewpoint of the whole system. Herein, this modeling approach has been utilized to reconstruct the mitochondrial metabolic network to integrate and analyze its capability of producing energy. Besides, extreme pathways analysis (EPA) and shadow prices analysis have also been integrated to study the interior characters of the network. Our modeling results have indicated for the first time that the covalent regulative property of pyruvate dehydrogenase is restrained by the feedback of acetyl-CoA. Combined with the biological experiments, these simulations <i>in silico</i> could be pretty useful for the further understanding of functions and characters of the biological network as a complex system.