基于WOA-BP算法的持液率预测模型研究

Prediction model of liquid holdup based on WOA-BP network

湿气管道在运行过程中,不可避免地会在管道低洼处出现积液。积液的存在会诱发很多安全问题,严重时甚至引发事故。因此,对湿气管道持液率进行预测就显得至关重要。文中基于灰色理论,对影响水平管道持液率的6个影响因素进行灰色关联分析,选取影响较大的因素作为影响变量;基于鲸鱼算法,建立鲸鱼算法优化BP神经网络的持液率预测模型,并与传统BP算法和遗传算法优化BP神经网络的预测模型进行对比。结果表明:文中预测模型的平均绝对误差为4.50%,均方根误差为0.0121,远低于传统BP预测模型和GA-BP预测模型,具有更高的计算精度,适用范围更广,为湿气管道的持液率预测提供了新思路和新途径。

During the operation of wet gas pipeline,it is inevitable that liquid will accumulate in the low-lying part of the pipeline.The existence of effusion will induce many safety problems,even lead to accidents.Therefore,it is very important to predict the liquid holdup of wet gas pipeline.In this paper,based on the grey theory,the grey correlation analysis of six factors affecting the liquid holdup of horizontal pipeline was carried out,and the more influential factors were selected as the influencing variables.Based on the whale algorithm,the liquid holdup prediction model of BP neural network optimized by whale optimization algorithm was established,and compared with the prediction model of BP neural network optimized by traditional BP algorithm and genetic algorithm.The results show that the average absolute percentage error of the prediction model is 4.50%,and the root mean square error is 0.0121,which is far lower than the traditional BP prediction model and GA-BP prediction model.It has higher calculation accuracy and wider application range,and provides a new idea and new way for the prediction of liquid holdup of wet gas pipeline.