Airborne Composition And Design Of The Fuel Measurements Embedded Fault-tolerant Computer Systems

Embedded system is a customized computer system with particular functions. It is widely applied in various fields due to its small size, low power consumption, low cost and high processing speed. There is far more applied amount of embedded system than of general-purpose computer. Embedded system has a widespread market perspective. On the other hand, in some particular fields (such as aerospace, geological survey, military application), the security and the reliability are two key issues upon the performance of embedded computer system. It beacomes especially important to research and apply embedded fault tolerance system.This paper describes the theory fundation, structural model, compositional method and evaluating procedure of fault tolerance computer system from a harware prospective, and designs hardware platform for board embedded fault tolerance computer system of airplane fuel measuring, which is based on CRTOS 2.0 real-time OS and Motorola PowerPC860 embedded processor, and modeled by a double-system hot standby prototype which is improved for practical applications. This involves the start and reset of main block of MPC860, memory assignment and management, response and processing of interrupt, external bus interface (1553B bus interface), peripheral communication interface (include RS-232, RS-422, RS-485, ARINC429), frequency counter interface, A/D converter and switch signal I/O controller, the arbitrage circuit for failure switching and fault tolerance handling etc.. Except for the analogue controlling and bus driving devices, we use FPGA to implement controlling logic in order to provide expansibility of application platform.For the high requirement of universality and reliability in which this design involed, and on the premise of implementing most of applicable functions, we consider some assuring measurements for guaranteeing the stability of system operation, which include self-diagnosis for system's basic status (self-test), system hardware operation checking, system bus out-of-time monitoring, system single-point failure tolerance and double-system switching on a board basis etc..