油田加热炉燃烧过程的神经网络模型辨识与自校正控制

加热炉在油田集输系统中有着广泛的应用。因运行时间长，其经济燃烧指标的高低直接影响着油田的生产成本。为实现加热炉出口参数的最佳调节及其经济燃烧，针对加热炉多变量、非线性、大滞后等特点，采用神经网络模型辨识的方法，建立了以加热炉为被控对象的神经网络正、逆模型，并且构成了神经网络内模自校正控制仿真系统。仿真研究表明，只要恰当地选择神经网络正、逆模型的结构和辨识数据的长度等参数，实现加热炉神经网络内模自校正控制的结果是令人满意的。

ANN Modeling and Self-tuning Control of the Oil Field Heating Furnace

The heating furnace is widely used in the oil field. Since the fact that the heating furnace needs a long period of running, the thermal efficiency affects oil field operation costs directly. For the purpose of improving the efficiency of the heating furnace, which is featured MIMO, nonlinear and delay, ANN is applied in its modeling. Its ANN model and inverse model and the simulation of its ANN Internal Model Self-tuning Control are built. Simulation shows that if chosen the appropriate ANN structure and training data quantity, its ANN internal model self-tuning control can be realized and the results can be acceptable.