该文将人工神经网的基本原理应用于过程控制系统，对辨识网络的基本结构以及网络的训练方法进行了初步探索和总结。在控制系统的辨识网络中应用基本ＢＰ算法与加动量项的ＢＰ算法、随机设置初值与训练设置初值、带参考模型与无参考模型对系统过渡过程特性的影响作了一定的比较与分析。仿真结果表明：所采用的改善网络自适应能力的方法，在不同程度上对系统的调节品质均有相应的改善。 This paper applies the basic principle of artificial neural network to process control system, and explores and summarizes the basic structure of identification network and the training method of network. In the control system identification network, the basic BP algorithm and the additive BP algorithm are used to set the initial value and the training initial value at random. The influence of the reference model and the non-reference model on the characteristics of the transition process are compared and analyzed. . The simulation results show that the methods used to improve the network adaptability have corresponding improvement on the adjustment quality of the system to varying degrees.