作为一个非线性、时变、多变量以及大时滞对象,直接空冷凝汽器很难建立起精确的数学模型,因此常规的控制方案效果不理想。利用BP神经网络建立直接空冷凝汽器的模型,并对于某一特定工况下,预测出环境温度、风速、负荷以及风机转速对直接空冷凝汽器背压的影响关系曲线。 As a non-linear, time-varying, multivariable and time-delay object, direct air-cooled condenser is difficult to establish an accurate mathematical model, so the conventional control scheme is not effective. The model of direct air-cooled condenser is established by using BP neural network and the influence curve of ambient temperature, wind speed, load and fan speed on backpressure of direct air-cooled condenser is predicted under a certain working condition.