Hot backup fault-tolerance method based on real-time operating systems and used for three satellite borne computers

Abstract

The invention discloses a hot backup fault-tolerance method based on real-time operating systems and for three satellite borne computers. The method comprises the following steps: firstly constructing three computers installed with the same real-time operating systems; in each control period, acquiring data of the other two computers by one single computer through data exchange among the three computers; comparing a first computer with a second computer, the first computer with a third computer and the second computer with the third computer; setting a mark indicating whether the first computer is in a good condition or not according to the consistence of data comparison results in combination with factors indicating whether each single computer is reset or not, whether ''a computer switching command word'' is valid or not and the like; running the three single computers synchronously in the same external output control states, and making one on-duty computer be responsible for final external output. During failure process, the three single computers run synchronously, state acquisition is not necessary during on-duty-computer switching, so that the failure recovery time is shortened, high instantaneity is achieved, gaps in system control are eliminated in a switching process, stable transition of system control can be realized, and independent reconfiguration of the systems is realized.

Description

technical field [0001] The invention relates to a fault-tolerant method for a three-machine redundant hot backup of an on-board computer. Background technique [0002] ERC32CPU (TSC695) is a dedicated chip for aerospace, with strong shock resistance and radiation resistance. The design idea and method of TSC695 real-time operating system are as follows: [0003] 1) The task scheduling adopts a priority-based preemptive scheduling strategy, and the task priority adopts a fixed priority; [0004] 2) Task state transition selection is performed among three task states (execution, ready, sleep); [0005] 3) Based on the premise that there are few external interrupt sources in the application system and no nesting occurs, the interrupt management program adopts the principle of not allowing external interrupt nesting to simplify the design; [0006] 4) Provide a mechanism for quick response to interrupts, so that the fastest response speed can be achieved when responding to so...

AI Technical Summary

Problems solved by technology

At present, there is no implementation method based on the three-machine hot backup fault-tolerant structure, so it cannot meet the actual needs of aerospace development.

Images