椭圆形覆冰导线的液滴碰撞特性

Research on Droplet Collision Characteristic of Elliptical Iced Conductor

过冷却液滴碰撞系数是架空输电线路覆冰生长机理研究的基础内容,该系数受覆冰导线截面形状、气象条件等因素的影响。为此以圆形、椭圆形覆冰为研究对象,采用流体力学数值模拟计算方法,分析结冰对象尺寸、风攻角、液滴中值体积直径对液滴碰撞系数的影响;随后,在大量液滴碰撞系数数据的基础上,采用误差反向传播(back propagation,BP)神经网络算法拟合出碰撞系数与其他因素的关系,在给定输入参数时可快速预测出液滴的碰撞系数;最后,搭建液滴碰撞系数实验平台。实验在0~3℃的环境温度条件下进行,通过高性能吸水纸吸附碰撞液滴,验证了数值模拟及BP神经网络预测过冷却液滴碰撞系数的有效性。

Collision coefficient of overcooling droplet which is affected by factors such as cross-section shapes, meteorological conditions is the basic research content of icing growth mechanism for the overhead power transmission line. This paper takes cylindrical and elliptical icing as research objects and uses the numerical simulation and calculation method based on computational fluid dynamics to analyse influence of sizes of the icing object, wind attack angles and median volumes of droplet on collision coefficients of droplet. On the basis of a large number of data of collision coefficient, it adopts the algorithm of error back propagation （BP） neural network to fit out relationship between collision coefficients and other factors, meanwhile, it can rapidly forecast collision coefficients of droplet as the input parameters are given. Finally, it builds a platform for testing collision coefficient and verifies effectiveness of this research method.