以离散 Hopfield神经网络为例 ,详细研究了学习算法对反馈神经网络故障性能的影响。简要介绍了离散Hopfield神经网络及其采用的 Hebb学习算法。详细分析了连接故障在学习过程中对网络连接权值故障性能的影响。给出了网络输出状态概率分布变化的计算公式。进行了计算机仿真 ,验证了理论分析结论的正确性 Taking discrete Hopfield neural network as an example, the effect of learning algorithm on the performance of feedback neural network is studied in detail. The discrete Hopfield neural network and its Hebb learning algorithm are briefly introduced. The effect of connection failure on the performance of network connection weights in learning process is analyzed in detail. The formula for calculating the change of probability distribution of network output is given. Computer simulation is conducted to verify the correctness of the theoretical analysis