基于IWOA-LSTM的水利设施裂缝数据预测OA
Prediction of Fracture Data of Water Conservancy Facilities Based on IWOA-LSTM
水利设施的裂缝宽度对水利设施的安全运行有着很大的影响,因此需要对裂缝数据预测.设计了一种基于改进鲸鱼优化算法(IWOA)与长短期记忆神经网络(LSTM)的组合网络模型(简称IWOA-LSTM),解决了普通的LSTM网络在预测水利设施裂缝数据时,人为设定超参数对预测精度的影响.该模型在现有鲸鱼优化算法的基础上,采用分段线性混沌映射初始化种群,并引入非线性收敛因子,加快种群收敛速度.结合某水电站4#号引水隧洞的裂缝数据,采用IWOA-LSTM模型进行预测,将预测结果与ARIMA模型、单一LSTM模型的预测结果进行对比,结果表明IWOA-LSTM模型的R方值分别提高了10.95%和4.69%,表明该模型有更高的预测精度.
The crack width of water conservancy facilities has a great impact on the safe operation of water conservancy facili-ties,so it is necessary to predict the crack data.In this paper,a combined network model(IWOA-LSTM for short)based on im-proved whale optimization algorithm(IWOA)and long short term memory network(LSTM)is designed,which solves the influence of artificial setting super parameters on the prediction accuracy when the ordinary LSTM network predicts the fracture data of water conservancy facilities.Based on the existing whale optimization algorithm,the model adopts piecewise linear chaotic map to initial-ize the population,and introduces a nonlinear convergence factor to accelerate the population convergence.Combined with the crack data of No.4 headrace tunnel of a hydropower station,IWOA-LSTM model is used for prediction.The prediction results are compared with those of ARIMA model and single LSTM model.The results show that the R value of IWOA-LSTM model is in-creased by 10.95%and 4.69%respectively,indicating that the model has higher prediction accuracy.
袁自祥;杨涛;皮明
西南科技大学信息工程学院 绵阳 621010||特殊环境机器人技术四川重点实验室 绵阳 621010西南科技大学信息工程学院 绵阳 621010||特殊环境机器人技术四川重点实验室 绵阳 621010西南科技大学信息工程学院 绵阳 621010||特殊环境机器人技术四川重点实验室 绵阳 621010
信息技术与安全科学
水利设施时间序列预测改进鲸鱼优化算法长短期记忆网络
water conservancy facilitiestime series predictionimproved whale optimization algorithmlong short term memory network
《计算机与数字工程》 2026 (3)
630-633,657,5
国家重点研发项目(编号:2019YFB1310504)资助.
评论