非线性过程的建模和控制问题尚无通用的模型结构和方法可用于过程控制。结合小波和神经网络方法进行过程控制的通用模型和方法的研究，提出了一种仅用尺度函数逼近的小波神经网络模型，并用它来实现非线性预测控制方案。该模型能用线性最小二乘方法进行拟合，因而具有易于实现和通用的优点。由于简化了在线优化方法，所提出的非线性预测控制方案已在线实现。用该方法进行了两个非线性系统的模型辨识和一个双线性系统的控制仿真，模型的通用性、辨识和控制方法的简单易用和仿真结果表明，所提出的方案对过程工业和非线性、高阶系统有很好的实用性，有着比标准ＰＩＤ控制器好得多的控制效果。 Modeling and Controlling Nonlinear Processes There is no universal model structure and method available for process control. Based on the research of general models and methods of process control based on wavelet and neural network, a wavelet neural network model approximated by scale function is proposed and used to realize nonlinear predictive control scheme. The model can be fit using linear least squares method, so it has the advantages of easy to implement and universal. Due to the simplified online optimization method, the proposed nonlinear predictive control scheme has been implemented online. The method is applied to the model identification of two nonlinear systems and the control simulation of a bilinear system. The generality of the model, the simple and easy-to-use method of identification and control and the simulation results show that the proposed scheme is effective for the process industry and the non- Linear, high-level systems have good usability and have much better control than standard PID controllers.