基于算术码实现了一种实时性较高的检错方法。这种检错方法不但能检测运算结果故障 ,还能够检测运算器故障、时序故障及时间故障。编码方法最终由四种特征组成 :模为 A的算术码 ,数据标志 ,时间标志 ,时序标志。为了在系统的处理过程中保持码字的特征 ,分析推导了编码运算器的实现算法和补偿常数的产生算法。最终通过校验变量的整体标志与预计算出的标志是否一致来实现故障的检测 Based on the arithmetic code to achieve a higher real-time error detection method. This method of error detection can not only detect the results of operational failures, but also to detect the operator fault, timing fault and time fault. The coding method is finally composed of four characteristics: the arithmetic code of the modulus A, the data flag, the time flag and the timing flag. In order to maintain the characteristics of the code words in the process of the system, the algorithm of the coder and the algorithm of the compensation constant are derived. Finally, through the verification of the overall sign of the variable and the estimated sign is consistent with the implementation of fault detection