Complex equipments often experience natural wear and external forces, and external forces cause a sudden increase in wear. In engineering practice, many high-reliability and long-lifetime equipments can not obtain a sufficient amount of failure data easily, it is necessary to use expert experience data to evaluate equipment's lifetime distribution and reliability indicators. However, the empirical data of experts is often not an exact constant, but a limited range. In order to deal with the uncertainty caused by human cognition, the uncertainty theory is introduced, and the empirical data of experts is described by uncertain variables. At the same time, the parameters in the distribution function are also uncertain and described by uncertain variables. The uncertainty theory is used to calculate the belief reliability of the equipment undergoing dependent competition failure process. When the parameters are uncertain variables, the belief reliability of the system is higher than that when the parameters are constant, The microelectronic mechanical system is used as an example to verify the effectiveness of the proposed model.