作者通过神经网络的任务分解在抽油泵故障诊断神经网络系统的研究 ,分析了神经网络任务分解的理论依据 ,并通过实验得出了神经网络任务分解时所需遵循的几个结论 ,以及通过实验证明神经网络任务分解对于减少网络的收敛迭代次数是一种行之有效的方法 The author studies the fault diagnosis neural network system of pumping oil pump based on the task decomposition of neural network and analyzes the theoretical basis of task decomposition of neural network. Some conclusions to be followed in the task decomposition of neural network are obtained through experiments. It is proved that the neural network task decomposition is an effective method to reduce the number of convergence iterations of the network