







摘" 要:為了推動水產品冷鏈物流行業高質量發展,對水產品冷鏈物流需求量的精準預測是實現水產品冷鏈行業快速發展和物流資源合理配置的基礎。針對目前冷鏈物流系統的復雜非線性,且統計數據樣本量少的特征,提出了一種基于BP神經網絡和支持向量機回歸的組合預測模型。文章從區域經濟、產品供給、冷鏈物流行業規模、社會四大維度選取10個指標構建影響因素指標體系,再結合各種預測方法的特點,選用BP-SVR組合預測模型。為驗證該組合模型的性能,以湖北省2002—2021的相關數據進行仿真預測。結果表明,該組合預測模型平均相對誤差僅為0.172,相比于單一的BP和SVR模型以及其他組合模型預測精度更高,因此使用BP-SVR組合預測模型能夠為湖北省未來水產品的需求量提供一定的參考價值。
關鍵詞:水產品;冷鏈物流;需求預測;BP-SVR組合模型
" 中圖分類號:F274" " 文獻標志碼:A" " DOI:10.13714/j.cnki.1002-3100.2024.15.037
Abstract: In order to promote the high-quality development of aquatic products cold chain logistics industry, accurate prediction of aquatic products cold chain logistics demand is the basis for realizing the rapid development of aquatic products cold chain industry and rational allocation of logistics resources. Aiming at the complex nonlinearity of the current cold chain logistics system and the small sample size of statistical data, a combined prediction model based on BP neural network and support vector machine regression is proposed. In this paper, ten indicators are selected from the four dimensions of regional economy, product supply, cold chain logistics industry scale, and society to construct the index system of influencing factors, and then combined with the characteristics of various prediction methods, the BP-SVR combined prediction model is selected. In order to verify the performance of the combined model, simulation prediction is carried out with the relevant data of Hubei Province from 2002 to 2021. The results show that the average relative error of the combined prediction model is only 0.172, which is higher than that of the single BP and SVR model and other combined models, so the use of the BP-SVR combined prediction model can provide a certain reference value for the future demand of aquatic products in Hubei Province.
Key words: aquatic products; cold chain logistics; demand forecast; BP-SVR combination model
0" 引" 言
水產品是冷鏈物流運輸重要對象之一。隨著居民對水產品需求量的增加,也給水產品冷鏈物流行業帶來了諸多挑戰。目前水產品冷鏈物流發展面臨著流通率低,運輸途中損耗高[1]等問題,為了加快水產品冷鏈物流的發展和保證水產品行業的供需平衡,對水產品冷鏈物流需求預測尤為重要。目前用于水產品冷鏈物流需求預測方法有灰色預測、BP神經網絡等[2-3]。但水產品冷鏈物流系統是一個復雜的非線性系統,水產品冷鏈物流的需求量往往受多重因素的影響。過往研究只用單一的預測模型進行研究,不能夠充分挖掘原始數據之間的潛在規律,存在一定的局限性,并未針對單一模型的預測優缺點進行組合創新,而組合模型在其他領域已經表現出較好的優越性[4-5]。……