|
| 摘要: |
| 汽包水位传感器信号的故障检测在热电厂生产过程中非常重要,通过分析主给水流量和主蒸汽流量的输出信号,建立预测水位输出的数学模型,提出了一种基于神经网络的汽包水位传感器故障检测方法,该方法首先建立神经网络结构;然后使用大量学习样本训练网络,确定权值和阈值;最后,进行系统仿真试验,比较仿真模型的输出与真实系统的输出,试验表明,用该方法进行汽包水位传感器的故障检测,效果令人满意。 |
| 关键词: 神经网络,故障检测,汽包水位,热电厂 |
| DOI： |
| 分类号:TP212 TP206.3 |
| 基金项目: |
|
| Fault detection for boiler water level sensor based on neural network |
|
CHEN Li-juan YANG Xin
|
| Abstract: |
| A model of predictive water level output is built up by analyzing the output signals of main water supply flow and main steam flow. A new method of fault detection for boiler water level sensor based on neural network is brought forward. The neural network is built up,and then trained with lots of learning samples to determine the weights and thresholds. The system is simulated and its output is compared with that of the real system. Results show its effectiveness. |
| Key words: neural network,fault detection,boiler water level,power plant |
