基于RBF网络的固相质量流量检测

气固两相流流动的复杂性和多样性直接影响了固相质量流量的测量精度,严重限制了其在工业生产中的应用。在双弯管法检测固相质量流量原理的基础上,利用人工神经网络优良的非线性映射能力,以两弯管处的压力差为输入,建立了一种基于径向基(RBF)网络的软测量模型,实现了对固相质量流量的在线测量。实验结果表明:该模型测量误差在3%以内,能够有效解决固相质量流量在线检测的问题,并且具有测量快速、精度高的优点。

Solid phase mass flow detection based on RBF network

Complexity and diversity of flow of gas-solid two-phase flow affect measurement precision of solid phase mass flow directly, limit its application in industrial production seriously. Based on principle of double-elbow method of solid phase mass flow detection, using powerful nonlinear mapping ability of artificial neural network, take pressure difference between the two elbows as input, a RBF-based network soft measurement model is established, realize online measurement of solid phase mass flow. Experimental results show that the model measurement error is within 3 %, can solve the problem of solid phase mass flow on-line detection effectively, and has advantages of fast, high precision.