Design Of Low Cost & Fault Tolerant Onboard Computing Unit Architecture

In recent years,with the continuous development of space technology,commercial space shows explosive growth.Based on the demand of commercial operation cost,orbiter payload and satellite are developing towards small constellation.As the core of satellite operation,the traditional onboard computing unit is mostly made of aerospace devices,which has a long development cycle.Aerospace devices have high reliability,but the cost is high,the range of options is small,easily affected by other factors,the supply is unstable and the performance is generally low.The commonly used commercial shelf products have the advantages of low price,wide range,mass supply and high performance,but the reliability of these devices is relatively low,and they are easily affected by high-energy particles in space environment.Therefore,how to use commercial shelf products to build a reliable onboard computing unit is an important problem.This topic starts from the reliability research of commercial shelf devices to build on-board computing unit,combined with the architecture and scheme of this kind of computing unit at home and abroad,uses the reliability model for modeling and quantitative analysis,takes the five-year task cycle as the predetermined design requirements,and obtains the applicable fault-tolerant architecture,which provides a theoretical basis for the subsequent hardware design.Secondly,due to the relatively low failure rate of some devices such as resistance,capacitance,power supply,logic gate and so on in the homogeneous processor module,the reliability of memory devices with high failure rate and prone to errors is strengthened.The Hamming code error detection and correction with information redundancy and the refresh of the backup cycle of stored information are used to realize the circuit correction of the possible errors of related devices Solid.Then,aiming at the possible faults of the same processor module in FPGA,the detection module,synchronization module and switching module are programmed to realize the fault-tolerant functions of data synchronization,fault monitoring and active standby switching.After that,the assembly and debugging of the onboard computing unit prototype are completed,and the test program is written.The fault tolerance function of the prototype is tested by hardware fault injection.Finally,the reliability of the prototype is modeled and analyzed.