Design Of Cube - Star High Reliability Spaceborne Computer Based On Dual - Mode Redundancy

With the space mission tend to be diversified, the cubesat which has the characteristics of low cost, high integration, short development cycle is becoming a new hotspot of aerospace research, and widely used in technical verification, to the ground observation, communication relay and so on. As the center of the cubesat, the design of on board computer(OBC) not only meet the low cost, easy expansion and other characteristics belongs to cubesat, but also to meet the basic reliability of a spacecraft.Based on the characteristics of cubesat, this paper analyses the design of cubesat electronic system.In addition, the paper introduces a series of fault tolerance techniques on hardware and software.Combined with the design case of other cubesat which has been on orbit, the paperproposesadesign methodfora OBConhigh reliability which is suit of the cubesat development. This method is based on the duplication redundancy, and takes cold backup mechanism. The paper design arbitration module and EDAC module in FPGA to prevent the total ionizing dose(TID) and single event effect(SEE), and realizes the detection of CPU switching and error detection and correction(EDAC) of memory.Considering the accumulation of SEE, the paper adds refresh mechanism into EDAC so as to improve the ability of preventing from SEE.In the matter ofsoftware fault-tolerant, Firstly, this paper focuses on the protection of critical data, procedure, function stack and proposes corresponding protect measures. Second, the paper builds the software voting system to avoid arithmetic error in running, and designs the task-level synchronization mechanism and expert voting system to solve the real-time problem and the short of single voting strategy. In addition, this paper design a dual backup bus system and realize the bus fault node localization and switching through establishing the node mutual monitoring mechanism.In order to solve the data transmission continuously whileswitching nodes, this paper a link oriented transport layer protocol is designed to ensure the reliability of data transmission.Finally, the paper evaluates the performance of prevent SEU through fault injection test. The results show that the system has a certain fault tolerance for single particle inversion.