Improved NSGA (non-dominated sorting genetic algorthm) -Ⅱ was used to optimize the network topology of aero-engine distributed control system while setting the reliability and weight of the control system as subjects. Taking the former reliability evaluation method's shortcoming into consideration

a method was proposed. Results show that reliability and weight of the control system increase with the increasing of nodes added into the circle topology or the increase of fundamental cycle's number. But when the number of fundamental cycle is larger than two

weight increases quickly without obvious increase of control system reliability. So

the number of fundamental cycles in aero-engine distributed control system topology shouldn't be larger than two in order to keep a balance between the topology's weight and the reliability.