过热汽温调节过程是一个非线性时变过程。目前广泛使用的ＰＩＤ控制虽然有结构简单、鲁棒性强的优点，但也存在自适应性能差的缺点。作者利用神经元的学习特性，给出了一种神经元网络控制模型，并将其与串级控制相结合，形成了过热汽温智能控制结构。该结构可以实现参数的自整定，具有较强的自学习能力。最后通过仿真验证了该智能控制的有效性。图５参６ Superheated steam temperature regulation process is a nonlinear time-varying process. Currently widely used PID control has the advantages of simple structure and strong robustness, but also has the disadvantage of poor adaptive performance. Using the learning characteristics of neurons, the author gives a neural network control model and combines it with cascade control to form a superheated steam temperature intelligent control structure. The structure can realize the self-tuning of parameters and has strong self-learning ability. Finally, the effectiveness of the intelligent control is verified by simulation. Figure 5 reference 6