基于BP神经网络的直接空冷凝汽器出口风温预测

Prediction of outlet air temperature of direct air condenser based on BP neural network

由于直接空冷系统的动态参数监测技术比较单一,为了便于对出口风温测点的研究,提出一种基于BP神经网络的空冷凝汽器出口风温预测模型。以某矸石电厂2×300 MW直接空冷凝汽器为对象进行研究,分别从锅炉侧与凝汽器侧分析影响空冷凝汽器出口风温的因素,从现场DCS系统分别采集各因素历史数据,对数据进行预处理后,运用BP神经网络建立空冷凝汽器出口风温预测模型。用现场实际运行数据验证模型的有效性。该模型可以预测不同工况下的空冷凝汽器的出口风温,为下一步制定动态参数监测技术提供依据,合理安排空冷凝汽器温度测点的位置和数量,也可以为空冷凝汽器的清洗周期做一个参考。

The dynamic parameter monitoring technology for direct air cooling system is relatively simple.In order to facilitate the research of outlet wind temperature measurement point,this paper proposes a BP neural network based air conditioning model for air condenser cooling.Taking a coal gangue power plant 2×300 MW direct air-cooled steam condenser as an object for research.The factors affecting the air temperature at the outlet of the air condenser are analyzed from the boiler side and the condenser side respectively.The historical data of each factor are collected from the on-site DCS system.After the data is preprocessed,the BP neural network is used to establish the air temperature prediction model of the air condenser.The actual operational data of the site is used to verify the validity of the model.The model can predict the outlet air temperature of the air condenser under different working conditions,provide a basis for the next step of developing dynamic parameter monitoring technology,and reasonably arrange the position and quantity of the air condenser temperature measuring point,or it can be an air condenser.The cleaning cycle is a reference.