【摘要】：借助多元統(tǒng)計方法識別2000年1月~2016年12月江西省生豬價格波動的成因,選擇生豬價格波動率作為警情指標,借助時差相關法,確定風險預警指標,建立BP人工神經(jīng)網(wǎng)絡模型,預警生豬價格波動的風險。結果表明,玉米價格、仔豬價格、豬肉價格、活雞價格、豆粕價格、生產(chǎn)者預期和疫情為主要警情指標,其中先行性指標是仔豬價格和生產(chǎn)者預期,同步指標是玉米價格、豬肉價格和活雞價格,豆粕價格和疫情是滯后指標;除個別樣本點外,BP神經(jīng)網(wǎng)絡模型輸出的生豬價格預警值和實際價格數(shù)據(jù)比較接近,生豬價格波動風險預警模型具有良好的預警效果。根據(jù)不同級別的風險程度提出相應對策建議,以便政府能夠根據(jù)不同警情采取具有針對性的措施,使生豬生產(chǎn)、市場平穩(wěn)健康發(fā)展。
[Abstract]:The causes of pig price fluctuation in Jiangxi Province from January 2000 to December 2016 were identified by means of multivariate statistical method. The hog price fluctuation rate was selected as the warning index and the risk early warning index was determined by means of time difference correlation method. The BP artificial neural network model was established. Risk of hog price fluctuations. The results showed that corn price, piglet price, pork price, live chicken price, soybean meal price, producer expectation and epidemic situation were the main warning indicators. Pork price and live chicken price, soybean meal price and epidemic situation are lagging indicators, except for a few sample points, the warning value of live pig price output by BP neural network model is close to the actual price data. The risk early warning model of hog price fluctuation has good early warning effect. According to different levels of risk, the corresponding countermeasures and suggestions are put forward, so that the government can take targeted measures according to different police situation, so as to make pig production and market develop smoothly and healthily.
【作者單位】： 江西農(nóng)業(yè)大學江西現(xiàn)代農(nóng)業(yè)發(fā)展協(xié)同創(chuàng)新中心;江西農(nóng)業(yè)大學理學院;
【基金】：國家自然科學基金“生豬價格波動的形成機理及風險預警仿真”(編號:71561014) 江西省社會科學項目“結構突變視角下江西農(nóng)產(chǎn)品價格市場形成機制研究”(編號:16YJ34) 江西高校人文社會科學基金“新常態(tài)下生豬價格的時空特征及波動機理研究”(編號:GL162014) 江西省教育廳項目“基于支持向量機和馬爾科夫模型的江西省生豬價格預警研究”(編號:GJJ160410)的階段性成果
【分類號】：F323.7;TP183

【相似文獻】

相關期刊論文 前1條

1 劉婕;張仙;周鋁;;多Agent的生豬價格預警模型研究[J];農(nóng)業(yè)網(wǎng)絡信息;2014年03期

相關碩士學位論文 前1條

1 曾弘;基于BP神經(jīng)網(wǎng)絡的生豬價格周期性預測[D];湖南農(nóng)業(yè)大學;2016年

，

本文編號：2262880