基于改进海洋捕食者算法优化GRU的短期风速预测  

Short term wind speed prediction based on improved MPA optimized GRU

在线阅读下载全文

作  者:李春梅 谷佳澄 王善求 谭佳伟 LI Chun-mei;GU Jia-cheng;WANG Shan-qiu;TAN Jia-wei(School of Mathematics and Statistics,Changchun University of Technology,Changchun 130012,China)

机构地区:[1]长春工业大学数学与统计学院,吉林长春130012

出  处:《白城师范学院学报》2024年第5期29-39,共11页Journal of Baicheng Normal University

基  金:吉林省科技发展计划项目(20230204078YY)。

摘  要:为了提高风速预测精度,提出了一种新的短期风速预测模型IMPA-GRU,其中IMPA算法是针对海洋捕食者算法初始化不均匀、收敛速度慢和收敛精度低等弱点引入Tent混沌映射、锦标赛选择策略和自适应阶段转换策略的一种改进算法.将IMPA算法与门控循环单元网络结合,通过算法优化网络参数,构建预测模型IMPA-GRU进行短期风速预测.为了评估模型性能,首先将IMPA算法与6种优化算法分别在8个测试函数上进行性能测试,然后将新提出的预测模型与MPA-GRU,GRU等5种模型进行对比.结果表明:IMPA算法在所有测试函数上都取得了更好的结果,IMPA-GRU模型在风速数据集上的MAPE,RMSE和MAE分别平均减小了0.98%,1.51%和1.58%,R2提高了0.56%,进一步提高了风速预测精度,减小了预测误差.In order to improve the accuracy of wind speed prediction,a new short-term wind speed prediction model IMPA-GRU is proposed.IMPA is an improved algorithm that introduces Tent chaotic mapping,tournament selection strategy,and adaptive stage transition strategy to address the weaknesses of MPA such as uneven initialization,slow convergence speed,and low convergence accuracy.Then,IMPA is combined with GRU to optimize GRU parameters,and IMPA-GRU is constructed for short-term wind speed prediction.In order to evaluate the performance of the model,firstly,IMPA and 6 comparative optimization algorithms were tested on 8 test functions,and then IMPA-GRU was compared with 5 predictive models such as MPA-GRU and GRU.The results showed that IMPA achieved better results on all test functions.The MAPE,RMSE,and MAE of IMPA-GRU on the wind speed dataset decreased by an average of 0.98%,1.51%,and 1.58%,respectively.R2 increased by 0.56%,further improving the accuracy of wind speed prediction and reducing prediction errors.

关 键 词:风速预测 海洋捕食者算法 门控循环单元 混沌映射 自适应选择 

分 类 号:TP18[自动化与计算机技术—控制理论与控制工程]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象