摘要
在液压系统中,液压管路是实现压力传导功能的重要组成部分,其压力值的变化不容忽视。在环境误差等因素的影响下,液压管路的压力变化呈现非线性和不稳定性。为解决该问题,提出基于粒子群优化算法(PSO)改进的基于注意力机制(Attention)的长短期记忆神经网络(LSTM)的液压管路压力预测方案。用某飞机液压管路的压力检测值作为输入数据,实现液压管路某支路位置的压力预测,并完成预测结果的可视化。实验结果表明,该模型预测平均误差为1.78%,符合液压管路压力预测要求。
Hydraulic pipeline,whose changes of pressure can't be ignored,is the most necessary component to transfer the pressrue in hydraulic system.The change of pressure at hydraulic pipeline,which is influenced by factors such as environmental errors,is non-linear and instability.In order to solve the problems,a method that pressure prediction based on PSO-LSTM-Attention is proposed.The research uses datas which drive from the actual pressure detection of a plane.PSO-LSTM-Attention model is used by the research on the pressrue prediction at the certain branch of hydraulic pipeline and the prediction of the model are visualized.The experinmental results indicates that average error of the model prediction is 1.78%,which meets the requirement of pressure prediction at hydraulic pipeline.
作者
李昂
徐梓敬
徐凯宏
谈子所
LI Ang;XU Zi-jing;XU Kai-hong;TAN Zi-suo(College of Computer and Control Engineering,Northeast Forestry University,Harbin 150040,China;College of Home and Art Design,Northeast Forestry University,Harbin 150040,China;Shanghai Wangyuan Instruments of Measurement Co.,Ltd.,Shanghai 200301,China)
出处
《中国电子科学研究院学报》
北大核心
2023年第12期1094-1099,共6页
Journal of China Academy of Electronics and Information Technology
基金
黑龙江重点研发计划(GZ20210017,GZ20210018,GZ20210019)。
关键词
液压管路
粒子群算法
LSTM
压力预测
注意力机制
hydraulic pipeline
particle swarm optimization algorithm
LSTM
pressure prediction
attention mechanism
