With the increasing application of multi axis synchronous control systems, the requirements of safety and reliability are becoming more and more urgent. Even when the original components are gradually degraded, many systems still have the expectation of continuous working hours. Thus, a dynamic matrix-PID cascade life extension control (DMC) method based on system residual life (RUL) prediction is proposed. Firstly, without changing the structure of the original multi axis synchronization system, a new architecture with real-time life prediction and life extension function is constructed by introducing the DMC as the life extension controller and the PHM module as the independent maintenance analysis decision-maker. Secondly, by establishing the relationship between the perceptual information such as the implicit degradation state of the actuator and the performance of the multi axis synchronization system, the analytical solution of the RUL distribution of the multi axis synchronization system is obtained and used for real-time prediction. Then, it is combined with the expected working time of the synchronization system as the cognitive basis of on-line autonomous maintenance, and based on an adaptive parameter adjustment law which depends on both actuator degradation state and life extension period, then used to dynamically adjust the element values of DMC constraint matrices Q and R on-line. The degradation of the actuator is slowed down and the service life of system is prolonged by better matching the characteristics of the degradation process and the demand for life extension, so as to achieve a better compromise balance between the performance and durability of the multi axis system. Finally, the effectiveness of the proposed method is verified by the simulation experiment of the stage multi motor synchronous control system.