Jupiter and beyond Deep space exploration missions are faced with engineering constraints such as large scale, long mission period and limited ground support, and it is urgent to solve the long-term fully autonomous flight control technology of deep space probes. In this paper, the autonomous navigation method based on optical measurement of spatial reference target and intelligent search algorithm of sequence target are used to realize the whole process of autonomous navigation and positioning. The online flight trajectory optimization technology based on autonomous navigation is designed to meet the online planning requirements of multi-target detection missions. An algorithm based on integrator measurement data fusion is proposed to realize all-state autonomous flight management and control. Through simulation verification, the autonomous navigation accuracy is better than 10-4ρ (ρ is the distance from the reference target), the trajectory planning detection target is not less than 5, the fault identification and early warning accuracy is higher than 90%, which can directly support the implementation of Jupiter, solar system marginal exploration and other engineering tasks.