气力喷头雾化特性的实验测定与神经网络数值模拟

在改变气力喷头的空气过流面积、吸液面与喷口之间的高度差和压力的条件下 ,测量了雾滴的直径 (体积中径DVMD) 用改进的BP算法人工神经网络对测试数据进行建模 ,通过建立好的模型进行仿真 从仿真结果发现 :在其余参数不变的情况下吸液高度对雾滴直径的影响呈线性递增关系 ;吸液高度不变时 ,对于不同的空气过流面积都存在使雾滴直径最小的压力值 ,并且都存在在低压区雾滴直径变化较小 ,在高压区雾滴直径变化较大的趋势 ;在吸液高度不变时 ,不同的空气过流面积都存在空气压力越高雾滴直径越细的现象

Experimental Determination of Atomization Characters of Pneumatic Nozzles and Parameters Simulation Based on Neural Network

The diameters of drops emitted from pneumatic atomizing nozzles are determined in different airflow areas,air pressures and sucked liquid height.A model is established by using an improved neural network based on BP algorithm,and the simulation results are given based on this model.From the results,some important characters are found:there is a linear and increasing relationship between sucked liquid height and drop diameters when other parameters keep constant;there is an air pressure producing minimum drop size for different airflow areas when sucked liquid height keeps constant;the higher the air pressure,the smaller the drop size is for different airflow areas when sucked height keeps constant.