基于WOA-BP组合模型的芦笋价格预测研究  

Asparagus Price Prediction Based on WOA-BP Combination Model

作　　者：杨洁 王俊美[1] 张超[1] YANG Jie;WANG Jun-mei;ZHANG Chao(College of Information Science and Engineering/Shandong Agricultural University,Tai'an 271018,China)

机构地区：[1]山东农业大学信息科学与工程学院,山东泰安271018

出　　处：《山东农业大学学报(自然科学版)》2025年第1期93-100,共8页Journal of Shandong Agricultural University:Natural Science Edition

基　　金：基于农业大数据芦笋价格数据库开发(381724);肥城人工智能机器人及智慧农业服务平台(381387)。

摘　　要：芦笋作为一种高价值蔬菜,价格走势预测对于市场分析和决策制定具有重要意义。芦笋价格受到多类因素的影响,因此提高价格预测精度的关键在于深入分析这些影响因素。本文提出了一种基于鲸鱼优化算法(WOA)与反向传播神经网络(BP)相结合的组合模型。研究中,本文首先采用主成分分析(PCA)对影响因素进行特征降维,随后将主成分分析后的多维特征集和经过数据融合的一维特征集分别输入优化前后的BP神经网络进行预测分析。通过对比分析不同输入下模型的预测性能,实验结果表明:经过WOA算法优化后的模型在预测效果上显著提升。具体而言,WOA-BP组合模型相较于传统的BP模型,在均方根误差(RMSE)上提高了2.431,平均绝对误差(MAE)提高了2.553,平均绝对百分比误差(MAPE)提高了5.606,决定系数(R^(2))提升了0.131。此外,WOA-BP-fusion模型与BP-fusion模型相比,RMSE提高了1.926,MAE提高了1.638,MAPE提高了5.539,R^(2)提高了0.101。结果表明,WOA-BP组合模型在进行数据融合后,能够更有效地捕捉输入特征与芦笋价格序列之间的关系,显著提高了预测精度,增强了模型的泛化能力和鲁棒性。WOA优化算法不仅提升了BP模型的预测精度,而且在数据融合过程中显著增强了模型对价格变动的反应能力。As a high-value vegetable,the price trend prediction of asparagus is of great significance for market analysis and decision making.Asparagus price is affected by many factors,so the key to improve the accuracy of price prediction is to deeply analyze these factors.Asparagus price is affected by many factors,so the key to improve the accuracy of price prediction is to deeply analyze these factors.n this paper,we propose a combined model based on the combination of Whale Optimization Algorithm(WOA)and Back Propagation Neural Network(BP).In this study,Principal Component Analysis(PCA)is firstly used to reduce the dimension of the influencing factors,and then the multidimensional feature set after principal component analysis and the one-dimensional feature set after data fusion are respectively input into the BP neural network before and after optimization for prediction analysis.By comparing and analyzing the prediction performance of the models under different inputs,the experimental results show that the model optimized by WOA algorithm significantly improves the prediction effect.Specifically,compared with the traditional BP model,the WOA-BP combined model has the Root Mean Square Error(RMSE)increased by 2.431,the Mean Absolute Error(MAE)increased by 2.553,the Mean Absolute Percentage Error(MAPE)increased by 5.606,and the Coefficient of Determination(R^(2))increased by 0.131.In addition,compared with the BP-fusion model,the WOA-BP-fusion model has RMSE increased by 1.926,MAE increased by 1.638,MAPE increased by 5.539,and R^(2) increased by 0.101.The results show that the WOA-BP combined model can more effectively capture the relationship between the input features and the asparagus price series after data fusion,significantly improve the prediction accuracy,and enhance the generalization ability and robustness of the model.The WOA optimization algorithm not only improves the prediction accuracy of the BP model,but also significantly enhances the responsiveness of the model to price changes in the data fusion pro

关 键 词：鲸鱼优化算法 组合模型 主成分分析 多源数据融合 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]