基于函数链神经网络的管道煤气流量计量系统

Measurement system for gas flow in the pipeline based on function chain neural network

在管道煤气计量系统测量中引入了管道煤气相对湿度修正,并对采用湿度传感器转换相对湿度信号,利用函数链神经网络对管道煤气工况温度下所对应的水蒸汽饱和压力进行了拟合,得到了基于函数链神经网络的管道煤气流量计量模型和在线计量系统,从而可以大大简化管道煤气流量计量软件,在流量计设计范围内可以快速准确地实现管道煤气流量实时在线计量。实际应用结果表明:该计量系统测量管道煤气流量误差为±0.7%。

Relative humidity modifying is introduced into measurement system for gas in the pipeline and a humidity sensor is used to tmnsforrne the signals of relative humidity. Moreover, on-line measurement system and model are gotten through a fitting formula about water vapour saturated pressure in the gas at different temperature based on function chain neural network, which would greatly simplify the software of measurement about gas flow in the pipeline and would quickly realize on line measurement about gas flow in the pipeline at the design range of flowmeter. The application results show that the measurement system error for gas flow in the pipeline is ± 0.7 %.