【正文】
毛精紡前紡工藝參數(shù)重要性的BP網(wǎng)絡(luò)定量評價法小二黑體,居中,3倍行距劉 貴收稿日期(黑體小五):20070310 修回日期:20070516(由編輯部填寫)基金項目(黑體小五):國家經(jīng)貿(mào)委創(chuàng)新項目(02CJ140501)基金名稱(基金編號)(楷體小五)作者簡介(黑體小五):劉貴(1983—),男,博士生。主要研究方向為毛精紡加工過程建模及其職能決策預(yù)報與控制。于偉東,通訊作者,Email:wdyu@。第一作者姓名(出生年—),性別,職稱,學(xué)歷。主要研究方向。通訊作者姓名,Email。,于偉東1,2四號楷體,居中,單倍行距(1. 東華大學(xué) 紡織材料與技術(shù)實驗室,上?!?01620。 武漢科技學(xué)院 紡織與材料學(xué)院,湖北 武漢 430073)小五號宋體,居中,單倍行距摘 要摘要寫作方法:請用第三人稱的語氣陳述該文研究目的(即為了……,或者針對……問題,)、過程、方法(即采用的手段和方法)、結(jié)果和結(jié)論(即研究得出的結(jié)論),重點是結(jié)果和結(jié)論,背景信息、基本概念及對文章的自我評價不應(yīng)出現(xiàn)在摘要中,要達到只看摘要而不必看文章就可理解全文主要內(nèi)容的程度;摘要字?jǐn)?shù)應(yīng)控制在200~300 字,英文要與中文相對應(yīng)。(小五黑體)在BP神經(jīng)網(wǎng)絡(luò)建模技術(shù)的基礎(chǔ)上,提出利用神經(jīng)網(wǎng)絡(luò)輸入層與輸出層之間的網(wǎng)絡(luò)權(quán)值及其分布來求各輸入?yún)?shù)重要程度的方法。將采集到的毛精紡企業(yè)前紡工藝參數(shù)運用BP神經(jīng)網(wǎng)絡(luò)分別建立了粗紗CV值和粗紗單重的預(yù)測模型。結(jié)果表明:所建模型的平均相對誤差都低于3%;采用樣本數(shù)據(jù)驗證,。對所建模型的網(wǎng)絡(luò)權(quán)重進行提取,分別計算出13個輸入?yún)?shù)對粗紗CV值和粗紗單重的重要性,挖掘出顯著而有效的參數(shù)。經(jīng)對比認為,BP網(wǎng)絡(luò)法比多元回歸顯著性分析(MRSA)更為精準(zhǔn),可用于對實際生產(chǎn)加工的預(yù)報和控制。(小五宋體)關(guān)鍵詞(小五黑體)毛精紡;前紡工藝參數(shù);模型;BP神經(jīng)網(wǎng)絡(luò);定量評價法(小五宋體)中圖分類號(小五黑體): TS (小五宋體) 文獻標(biāo)志碼(小五黑體): Quantitative evaluation method for the significance of worstedforespinning parameters based on BP neural network小四Times New Rome,3倍行距LIU Gui1,YU Weidong1,2(五號)( Materials andTechnology Laboratory Donghua University, Shanghai 201620,China。 of Textiles and Materials, Wuhan University of Science and Engineering, Wuhan, Hubei 430073,China)小五,居中 Abstract Based on BP neural network model technology, a new approach was developed and applied to appraise the input parameters′significant degree through the weightiness and its distribution between the input and output layer. Usingthe forespinningworking procedure data gathered fromtheworsted textiles enterprise, the roving unevenness and weight prediction models were established respectively. The results indicated that the models′mean relative errors are all less than 3%。 the correlation coefficientR2between the prediction value and the actual are all more than 095. Using the weightiness extracted fromthe established models, the 13 input parameters′significance to the roving unevenness and weight were calculated respectively, and the remarkable and effective parameters are excavated out. Meanwhile contrasting to the multivariate regression significance analysis (MRSA), the BP neural network method is more exact than MRSA and can be used in the forecast and control of the actual produce and manufacture (小五)Key words Double glow。 Artificial neural network。 Prediction model (