于華竟，徐大明，羅 娜，邢 斌，孫傳恒

雜糧供應(yīng)鏈區(qū)塊鏈多鏈追溯監(jiān)管模型設(shè)計

于華竟，徐大明，羅 娜，邢 斌，孫傳恒※

（1. 國家農(nóng)業(yè)信息化工程技術(shù)研究中心，北京 100097；2. 農(nóng)產(chǎn)品質(zhì)量安全追溯技術(shù)及應(yīng)用國家工程實驗室，北京 100097）

針對雜糧產(chǎn)品供應(yīng)鏈鏈條長、主體多、區(qū)塊鏈追溯過程數(shù)據(jù)無法差異化共享、鏈上數(shù)據(jù)難以實時監(jiān)管等難題，通過分析雜糧供應(yīng)鏈環(huán)節(jié)業(yè)務(wù)流程與監(jiān)管特性，提出了基于區(qū)塊鏈多鏈架構(gòu)的雜糧追溯模型，并在此基礎(chǔ)上建立多鏈數(shù)據(jù)存儲架構(gòu)，設(shè)計了基于監(jiān)管授權(quán)組網(wǎng)建鏈的網(wǎng)絡(luò)準(zhǔn)入機制，并通過智能合約實現(xiàn)數(shù)據(jù)的鏈前監(jiān)管與追溯節(jié)點的鏈上管控。為驗證模型有效性，基于Hyperledger Fabric設(shè)計并實現(xiàn)區(qū)塊鏈追溯系統(tǒng)，對山西忻州雜糧應(yīng)用案例進(jìn)行分析。在安全方面，企業(yè)組網(wǎng)授權(quán)擴散性測試密文平均改變率為82.53%，相關(guān)性測試密文平均改變率為82.39%，具備較高的安全性與混淆性。在效率方面，消費者查詢公開追溯數(shù)據(jù)平均時間為0.415 s，監(jiān)管部門調(diào)用跨鏈接口查詢企業(yè)敏感追溯數(shù)據(jù)平均時間為0.871 s。結(jié)果表明，該研究設(shè)計并實現(xiàn)的面向監(jiān)管的雜糧多鏈追溯系統(tǒng)在滿足消費者追溯需求的基礎(chǔ)上，能夠?qū)崿F(xiàn)追溯數(shù)據(jù)賬本與鏈間交易記錄的實時管控，為農(nóng)產(chǎn)品區(qū)塊鏈追溯監(jiān)管系統(tǒng)研究提供借鑒與參考。

農(nóng)產(chǎn)品；追溯；區(qū)塊鏈；多鏈；數(shù)據(jù)監(jiān)管；Hyperledger Fabric

0 引 言

雜糧具有豐富的營養(yǎng)物質(zhì)，是中國西部地區(qū)農(nóng)業(yè)經(jīng)濟的主要來源[1]。但近幾年，食品質(zhì)量安全問題頻發(fā)，化學(xué)藥品濫用、重金屬超標(biāo)等嚴(yán)重影響人們對農(nóng)戶的信任[2]，農(nóng)產(chǎn)品信息亟須有效監(jiān)管。構(gòu)建雜糧追溯系統(tǒng)實現(xiàn)多方高效協(xié)作、快速追溯產(chǎn)品來源，能夠重筑雜糧與消費者間信任橋梁[3]?，F(xiàn)有的追溯系統(tǒng)研究集中于條形碼[4]、二維碼[5]、射頻識別[6]以及無線傳感網(wǎng)絡(luò)等記錄追溯數(shù)據(jù)，但中國雜糧存在種植、加工、物流等追溯鏈條長、生產(chǎn)分散、上下游信息斷鏈和數(shù)據(jù)不透明等問題[7]，傳統(tǒng)追溯系統(tǒng)難以關(guān)聯(lián)并且高效監(jiān)管各環(huán)節(jié)追溯記錄，出現(xiàn)質(zhì)量安全問題時無法及時召回問題產(chǎn)品、準(zhǔn)確定位責(zé)任主體、精準(zhǔn)管控追溯企業(yè)等問題。

區(qū)塊鏈鏈?zhǔn)浇Y(jié)構(gòu)具有數(shù)據(jù)不可篡改、正向記錄逆向追溯特點[8]，實現(xiàn)全維度、透明化追蹤記錄雜糧作物生產(chǎn)收獲全過程，為解決雜糧供應(yīng)鏈企業(yè)數(shù)據(jù)共享、數(shù)據(jù)價值深度挖掘等提供技術(shù)支撐[9]。近些年，國內(nèi)外學(xué)者從不同角度探究區(qū)塊鏈技術(shù)在農(nóng)產(chǎn)品追溯中的應(yīng)用，如楊信廷等[10]、于合龍等[11]、Yang等[12]采用鏈下中心化數(shù)據(jù)庫協(xié)同存證的方式探究區(qū)塊鏈技術(shù)在農(nóng)產(chǎn)品供應(yīng)鏈大數(shù)據(jù)中的應(yīng)用；李宣等[13]、Cao等[14]、Ding等[15]、李夢琪等[16]設(shè)計雙區(qū)塊鏈架構(gòu)存儲產(chǎn)品追溯海量數(shù)據(jù)。然而，由于雜糧供應(yīng)鏈追溯企業(yè)分散以及區(qū)塊鏈網(wǎng)絡(luò)賬本公開透明，具備商業(yè)價值的敏感數(shù)據(jù)隱私保護(hù)以及授權(quán)訪問等還需要進(jìn)行深入研究。此外，陳純[17]、洪學(xué)海等[18]指出區(qū)塊鏈網(wǎng)絡(luò)健康可持續(xù)的發(fā)展需要監(jiān)管技術(shù)的支持，區(qū)塊鏈安全監(jiān)管技術(shù)已經(jīng)成為區(qū)塊鏈規(guī)?；瘧?yīng)用的重點研究方向。在區(qū)塊鏈農(nóng)產(chǎn)品追溯系統(tǒng)數(shù)據(jù)監(jiān)管方面，張燕等[19]、王毅[20]、許繼平等[21]從追溯數(shù)據(jù)監(jiān)管方面探究區(qū)塊鏈技術(shù)在農(nóng)產(chǎn)品追溯中的應(yīng)用，霍紅等[22]從供應(yīng)鏈視角構(gòu)建區(qū)塊鏈農(nóng)產(chǎn)品質(zhì)量安全監(jiān)管體系，指出供應(yīng)鏈監(jiān)管還包括參與主體的行為、意識的管控。針對以上問題，在保證雜糧供應(yīng)鏈各企業(yè)間追溯數(shù)據(jù)隱私保護(hù)的基礎(chǔ)上，如何實現(xiàn)數(shù)據(jù)、交易行為的穿透式監(jiān)管，不僅體現(xiàn)在追溯數(shù)據(jù)的全方面監(jiān)管，也體現(xiàn)在對追溯企業(yè)節(jié)點的管控。因此，設(shè)計適用于雜糧供應(yīng)鏈的可監(jiān)管追溯具有重要意義。

本文通過分析雜糧供應(yīng)鏈全生命周期追溯流程，設(shè)計并實現(xiàn)基于區(qū)塊鏈多鏈架構(gòu)的可監(jiān)管雜糧追溯系統(tǒng)，并根據(jù)實際生產(chǎn)情況提出面向監(jiān)管的雜糧區(qū)塊鏈多鏈數(shù)據(jù)存儲模型，并設(shè)計基于監(jiān)管授權(quán)的網(wǎng)絡(luò)準(zhǔn)入機制，通過智能合約實現(xiàn)鏈上鏈下協(xié)同監(jiān)管，充分利用區(qū)塊鏈技術(shù)保障各類生產(chǎn)要素資源與雜糧實體經(jīng)濟深度對接，將雜糧供應(yīng)鏈與區(qū)塊鏈雙鏈融合，解決雜糧供應(yīng)鏈存在的數(shù)據(jù)擴容、數(shù)據(jù)共享、監(jiān)管難等問題。

1 雜糧供應(yīng)鏈關(guān)鍵信息分析

雜糧供應(yīng)鏈包括種植、加工、倉儲、運輸、銷售等多環(huán)節(jié)，由攝像頭、傳感器、北斗定位裝置等物聯(lián)網(wǎng)設(shè)備實時采集視頻、音頻、圖像、位置等產(chǎn)品操作信息，達(dá)到物理級真實，保證高品質(zhì)雜糧產(chǎn)品源頭安全。但在規(guī)?；瘧?yīng)用的雜糧產(chǎn)品追溯中，數(shù)據(jù)多源、海量、異構(gòu)、種類繁雜，其中并非所有數(shù)據(jù)用于共享追溯，涉及生產(chǎn)數(shù)量、成本等與產(chǎn)品質(zhì)量安全無關(guān)的敏感信息無法完全透明公開，并且供應(yīng)鏈主體間關(guān)于數(shù)據(jù)采集、存儲格式缺乏共識性標(biāo)準(zhǔn)，導(dǎo)致上下游企業(yè)間數(shù)據(jù)隱私性差、監(jiān)管部門無法高效監(jiān)管供應(yīng)鏈數(shù)據(jù)。針對上述問題，在保證產(chǎn)品基本信息逆向追溯的基礎(chǔ)上，為了提升監(jiān)管部門工作效率以及保證企業(yè)敏感數(shù)據(jù)的安全，本文提取供應(yīng)鏈各環(huán)節(jié)關(guān)鍵追溯數(shù)據(jù)并將其劃分為可公開共享數(shù)據(jù)與敏感數(shù)據(jù)，具體如表1所示。

表1 雜糧供應(yīng)鏈關(guān)鍵信息

2 基于區(qū)塊鏈多鏈的雜糧供應(yīng)鏈追溯模型設(shè)計

2.1 雜糧追溯區(qū)塊鏈多鏈架構(gòu)

中國雜糧供應(yīng)鏈追溯環(huán)節(jié)多、追溯數(shù)據(jù)海量異構(gòu)，而傳統(tǒng)追溯系統(tǒng)通過企業(yè)中心化數(shù)據(jù)庫存儲追溯數(shù)據(jù)，造成企業(yè)間數(shù)據(jù)流通即時性差并且存在篡改風(fēng)險[23]。區(qū)塊鏈以分布式賬本代替中心數(shù)據(jù)庫存儲企業(yè)數(shù)據(jù)，正向記錄產(chǎn)品信息、逆向追溯，實現(xiàn)供應(yīng)鏈上下游企業(yè)數(shù)據(jù)共享、價值傳遞，解決因企業(yè)間信任問題引起的數(shù)據(jù)壁壘問題。區(qū)塊鏈追溯節(jié)點通過區(qū)塊高度（Block height）或交易哈希（Transaction hash）等索引查詢數(shù)據(jù)賬本，無法做到鏈上數(shù)據(jù)的差異化共享[24]，而且區(qū)塊單鏈?zhǔn)酱鎯Y(jié)構(gòu)能夠保證區(qū)塊數(shù)據(jù)不可篡改、不可刪除的永久可查、可用，但面臨追溯大數(shù)據(jù)的擴容難題。在區(qū)塊鏈追溯規(guī)模化應(yīng)用后，因網(wǎng)絡(luò)節(jié)點共識串行處理數(shù)據(jù)交易而無法獲得傳統(tǒng)中心化追溯系統(tǒng)的性能表現(xiàn)[25]。

針對上述問題，根據(jù)雜糧實際應(yīng)用環(huán)境，綜合考慮監(jiān)管部門與供應(yīng)鏈環(huán)節(jié)，在生產(chǎn)、加工、倉儲、物流以及經(jīng)銷等供應(yīng)鏈企業(yè)獲取監(jiān)管部門組網(wǎng)授權(quán)文件后建立企業(yè)區(qū)塊鏈，并且由各企業(yè)追溯節(jié)點、消費者節(jié)點與監(jiān)管節(jié)點建立追溯區(qū)塊鏈，如圖1所示。通過區(qū)塊鏈通道技術(shù)建立面向企業(yè)數(shù)據(jù)存儲的生產(chǎn)鏈、加工鏈、倉儲鏈、物流鏈、分銷鏈五條企業(yè)區(qū)塊鏈，建立面向追溯需求、鏈間交易與可控監(jiān)管的追溯鏈，以企業(yè)鏈多鏈結(jié)構(gòu)和追溯主鏈的形式共同組建雜糧區(qū)塊鏈追溯系統(tǒng)，實現(xiàn)數(shù)據(jù)容量擴增、數(shù)據(jù)隔離存儲、消費者逆向追溯、鏈間交易留痕可控。追溯鏈作為區(qū)塊鏈追溯網(wǎng)絡(luò)主鏈，提供鏈間交易環(huán)境，實現(xiàn)雜糧供應(yīng)鏈多企業(yè)鏈間數(shù)據(jù)交互以及上下游企業(yè)間交易留痕管控。

為保證區(qū)塊鏈追溯網(wǎng)絡(luò)健康可持續(xù)穩(wěn)定發(fā)展，供應(yīng)鏈企業(yè)首先獲取監(jiān)管部門審核授權(quán)文件，利用通道技術(shù)建立企業(yè)追溯多鏈以及企業(yè)節(jié)點組建追溯網(wǎng)絡(luò)，通過通道天然的隔離性保證企業(yè)間數(shù)據(jù)的隱私性、安全性。監(jiān)管部門與各企業(yè)追溯節(jié)點加入追溯通道創(chuàng)建追溯主鏈，企業(yè)產(chǎn)品生產(chǎn)加工過程中涉及商業(yè)價值敏感數(shù)據(jù)存儲在企業(yè)鏈，具備追溯價值的可公開產(chǎn)品信息存儲在追溯主鏈。同時由于通道隔離上下游企業(yè)鏈間交易無法直接進(jìn)行，必須通過追溯主鏈交互完成信息交換與追溯價值傳遞。企業(yè)追溯節(jié)點在追溯主鏈上通過智能合約向監(jiān)管節(jié)點提供同構(gòu)多鏈跨鏈數(shù)據(jù)查詢接口，企業(yè)敏感追溯數(shù)據(jù)面向監(jiān)管追溯開放；同時對消費者節(jié)點提供追溯鏈上賬本查詢接口，公開追溯數(shù)據(jù)面向消費追溯開放。

2.2 追溯數(shù)據(jù)存儲與監(jiān)管模型

本文在深入分析雜糧供應(yīng)鏈特點的基礎(chǔ)上，基于區(qū)塊鏈技術(shù)建立面向監(jiān)管的雜糧多鏈追溯數(shù)據(jù)存儲模型，如圖 2所示。追溯數(shù)據(jù)按照供應(yīng)鏈環(huán)節(jié)從生產(chǎn)鏈開始記錄并通過追溯鏈共享直到產(chǎn)品消費，產(chǎn)品敏感信息由企業(yè)鏈加密存儲授權(quán)訪問、公開信息由追溯鏈透明共享、產(chǎn)品交易由追溯鏈留痕管控。其中，企業(yè)鏈通過追溯鏈交互共享數(shù)據(jù)，數(shù)據(jù)上鏈時首先調(diào)用監(jiān)管合約執(zhí)行數(shù)據(jù)鏈前監(jiān)管，判斷數(shù)據(jù)格式、內(nèi)容是否符合合約追溯條例；然后通過上鏈智能合約在企業(yè)節(jié)點共識記賬，更新區(qū)塊鏈企業(yè)節(jié)點數(shù)據(jù)庫與區(qū)塊索引記錄；最后觸發(fā)跨鏈合約將可公開共享追溯數(shù)據(jù)上傳追溯鏈。上鏈鍵值對追溯數(shù)據(jù)如表2所示，由供應(yīng)鏈企業(yè)物聯(lián)網(wǎng)設(shè)備實時采集和上游企業(yè)交易的產(chǎn)品數(shù)據(jù)存入?yún)^(qū)塊體默克爾樹葉子節(jié)點出塊記賬，并且以鍵值對的形式更新世界狀態(tài)數(shù)據(jù)庫，同時將返回的交易哈希、區(qū)塊高度存入世界狀態(tài)索引數(shù)據(jù)庫，通過區(qū)塊高度或交易哈希索引查詢追溯數(shù)據(jù)，降低查詢時間。

從數(shù)據(jù)存儲、數(shù)據(jù)監(jiān)管兩方面分析數(shù)據(jù)存儲模型。數(shù)據(jù)存儲中追溯文本數(shù)據(jù)選用CouchDB狀態(tài)數(shù)據(jù)庫以鍵值對方式存儲，通過鏈下索引數(shù)據(jù)庫檢索數(shù)據(jù)提升數(shù)據(jù)上傳與查詢效率，多媒體流數(shù)據(jù)通過星際文件系統(tǒng)（Interplanetary File System，IPFS）分布式存儲[26]。數(shù)據(jù)監(jiān)管一方面是鏈前數(shù)據(jù)與鏈上賬本的監(jiān)管，數(shù)據(jù)鏈前監(jiān)管是通過智能合約審計數(shù)據(jù)格式、內(nèi)容，保證數(shù)據(jù)上鏈時安全可靠；鏈上監(jiān)管是由監(jiān)管節(jié)點實時接收追溯鏈上廣播信息，根據(jù)動態(tài)關(guān)鍵字詞庫審計追溯數(shù)據(jù)賬本，出現(xiàn)違規(guī)追溯記錄時及時進(jìn)行數(shù)據(jù)取證、認(rèn)定；另一方面是追溯節(jié)點的管控，監(jiān)管部門根據(jù)審計實時反饋通過取消監(jiān)管授權(quán)的方式限制追溯節(jié)點鏈上數(shù)據(jù)賬本操作權(quán)限，為追溯網(wǎng)絡(luò)數(shù)據(jù)內(nèi)容安全提供有力支撐。

表2 上鏈鍵值對數(shù)據(jù)

2.3 企業(yè)組網(wǎng)建鏈與授權(quán)管控模型

區(qū)塊鏈基于P2P協(xié)議構(gòu)建去中心化系統(tǒng)，全網(wǎng)節(jié)點平等、數(shù)據(jù)無成本快速傳遞[27-28]。在雜糧追溯網(wǎng)絡(luò)規(guī)?；瘧?yīng)用后，受到上鏈數(shù)據(jù)量大、時效性高限制，在加強鏈前數(shù)據(jù)審查的基礎(chǔ)上，強化鏈上節(jié)點權(quán)限管控力度具有重要的研究意義。針對追溯節(jié)點監(jiān)管，本文設(shè)計基于監(jiān)管授權(quán)的網(wǎng)絡(luò)準(zhǔn)入機制，如圖3所示。供應(yīng)鏈企業(yè)通過非對稱國密算法SM2生成公私密鑰對，加入?yún)^(qū)塊鏈追溯網(wǎng)絡(luò)時，監(jiān)管部門獲取由供應(yīng)鏈企業(yè)廣播的公鑰，并接收由供應(yīng)鏈企業(yè)提交的授權(quán)申請；根據(jù)授權(quán)申請，對供應(yīng)鏈企業(yè)進(jìn)行建鏈資質(zhì)審核；在建鏈資質(zhì)審核通過的情況下，利用公鑰對授權(quán)文件進(jìn)行加密，生成授權(quán)密文；將授權(quán)密文作為建鏈資質(zhì)審核結(jié)果發(fā)送給供應(yīng)鏈企業(yè)，以供供應(yīng)鏈企業(yè)利用私鑰對授權(quán)密文進(jìn)行解密，并根據(jù)解密后獲取的授權(quán)文件構(gòu)建相應(yīng)的企業(yè)私有鏈；私鑰與公鑰相對應(yīng)。本文設(shè)計的追溯企業(yè)組網(wǎng)建鏈與授權(quán)管控模型，監(jiān)管部門通過對供應(yīng)鏈企業(yè)進(jìn)行建鏈?zhǔn)跈?quán)的方式，管控鏈上追溯節(jié)點的網(wǎng)絡(luò)賬本操作權(quán)限。

由于供應(yīng)鏈企業(yè)組網(wǎng)建鏈以及操作鏈上賬本數(shù)據(jù)都需要企業(yè)授權(quán)文件Key，所以追溯網(wǎng)絡(luò)的安全性取決于Key是否安全不可破解。監(jiān)管部門根據(jù)企業(yè)建鏈資質(zhì)生成授權(quán)密文，并通過哈希函數(shù)生成唯一不可逆的授權(quán)簽名，加密算法如公式（1）～（4）所示。為保證授權(quán)文件傳輸安全，如公式（5）～（6）所示通過企業(yè)公鑰加密傳輸、私鑰解密獲取授權(quán)建鏈組建企業(yè)追溯網(wǎng)絡(luò)。

根據(jù)授權(quán)過期時間信息、授權(quán)名稱信息、授權(quán)企業(yè)信息和簽發(fā)時間信息，生成授權(quán)文件的中間編碼密文Info，記為

監(jiān)管部門接收到供應(yīng)鏈企業(yè)的公鑰和授權(quán)申請后，生成授權(quán)文件密文并通過公鑰加密傳輸，生成具有保密性的的授權(quán)密文，由于只有企業(yè)保留的私鑰可以對授權(quán)密文進(jìn)行解密，進(jìn)而讀取解密后的內(nèi)容，可以有效防止信息泄露。記為

2.4 智能合約設(shè)計

智能合約在觸發(fā)條件滿足時可自動執(zhí)行合約邏輯，以信息化的方式傳播、驗證或執(zhí)行合約協(xié)議，使區(qū)塊鏈及時對外部治理做出反應(yīng)[29-30]。本文設(shè)計監(jiān)管智能合約，實現(xiàn)監(jiān)管自動執(zhí)行，強化數(shù)據(jù)鏈前監(jiān)管、鏈上安全，減緩監(jiān)管部門負(fù)載壓力。監(jiān)管合約將監(jiān)管法律條例轉(zhuǎn)化為合約條款，由雜糧供應(yīng)鏈企業(yè)主體參與協(xié)商制定，內(nèi)容公開透明且不可篡改，提供鏈上協(xié)作的基石，實現(xiàn)區(qū)塊鏈追溯節(jié)點組網(wǎng)授權(quán)管控、追溯數(shù)據(jù)鏈前與鏈上交易監(jiān)管、跨鏈數(shù)據(jù)穿透式查詢監(jiān)管，具體智能合約設(shè)計如表3。

本文通過聯(lián)盟鏈通道技術(shù)創(chuàng)建追溯多鏈網(wǎng)絡(luò)，為多條企業(yè)鏈和追溯鏈封裝出賬本數(shù)據(jù)交互接口，通過合約調(diào)用接口實現(xiàn)鏈間數(shù)據(jù)交互。下面介紹數(shù)據(jù)上鏈與查詢的智能合約算法，算法1描述追溯數(shù)據(jù)上鏈操作，算法2描述監(jiān)管追溯跨鏈查詢步驟，第7～9步、12～14中查詢追溯鏈公開賬本數(shù)據(jù)，面向消費者掃碼追溯可返回相應(yīng)企業(yè)公開追溯數(shù)據(jù)，面向監(jiān)管追溯則返回上游企業(yè)鏈追溯碼；第3～5步、15～17中通過智能合約調(diào)用企業(yè)鏈賬本查詢接口獲取企業(yè)鏈敏感追溯數(shù)據(jù)。

表3 智能合約設(shè)計

算法1：數(shù)據(jù)上鏈智能合約算法

輸入：企業(yè)追溯節(jié)點cTraPeer，企業(yè)鏈?zhǔn)跈?quán)文件Keyproc，追溯鏈?zhǔn)跈?quán)文件Keytra，追溯數(shù)據(jù)traceData

輸出：交易哈希txID，區(qū)塊高度blockNum，公開追溯數(shù)據(jù)traceData’

算法2：數(shù)據(jù)查詢智能合約算法

輸入：監(jiān)管節(jié)點sTraPeer，五條企業(yè)鏈?zhǔn)跈?quán)文件Key，追溯鏈?zhǔn)跈?quán)文件Keytra，產(chǎn)品追溯碼traceCode

輸出：企業(yè)鏈產(chǎn)品敏感追溯數(shù)據(jù)

3 面向監(jiān)管的雜糧區(qū)塊鏈多鏈追溯系統(tǒng)實現(xiàn)

3.1 系統(tǒng)架構(gòu)

通過聯(lián)盟鏈構(gòu)建雜糧產(chǎn)品追溯系統(tǒng)，基于通道技術(shù)創(chuàng)建多條企業(yè)鏈保證敏感數(shù)據(jù)的隱私保護(hù)與授權(quán)訪問，通過追溯鏈保證公開數(shù)據(jù)多方共享與價值傳遞，由監(jiān)管合約保證數(shù)據(jù)監(jiān)管條例的實時運行與鏈上數(shù)據(jù)賬本與節(jié)點的有效管控，實現(xiàn)追溯系統(tǒng)數(shù)據(jù)安全。

追溯系統(tǒng)架構(gòu)分為4層：應(yīng)用層、接口層、數(shù)據(jù)層、存儲層。其中，應(yīng)用層通過溫濕度和光敏等傳感器、北斗定位裝置、攝像頭等物聯(lián)網(wǎng)設(shè)備采集鏈下追溯數(shù)據(jù)確保數(shù)據(jù)源頭真實可信；面向消費者提供追溯掃碼查詢產(chǎn)品追溯信息，面向企業(yè)提供追溯數(shù)據(jù)跨鏈存儲以及追溯鏈上交易留痕可管控；面向監(jiān)管部門提供追溯企業(yè)節(jié)點授權(quán)管控和追溯數(shù)據(jù)穿透式監(jiān)管。

接口層封裝企業(yè)多鏈和追溯鏈賬本數(shù)據(jù)寫入與查詢操作接口，對監(jiān)管部門提供數(shù)據(jù)跨鏈操作接口和企業(yè)授權(quán)管控接口，對消費者提供追溯數(shù)據(jù)查詢接口。

數(shù)據(jù)層中企業(yè)節(jié)點授權(quán)監(jiān)管部門可穿透式監(jiān)管企業(yè)鏈敏感數(shù)據(jù)；授權(quán)消費者透明查詢公開追溯數(shù)據(jù)。消費者掃碼發(fā)送產(chǎn)品追溯碼查詢追溯鏈上數(shù)據(jù)賬本獲取追溯信息，監(jiān)管部門通過企業(yè)賬本操作授權(quán)以及由追溯鏈獲取的產(chǎn)品追溯碼跨鏈監(jiān)管企業(yè)敏感數(shù)據(jù)。

存儲層中區(qū)塊鏈數(shù)據(jù)賬本由文件系統(tǒng)存儲的區(qū)塊結(jié)構(gòu)和數(shù)據(jù)庫維護(hù)的世界狀態(tài)組成，通過智能合約執(zhí)行交易更新區(qū)塊鏈賬本。文件系統(tǒng)以默克爾樹哈希存儲上鏈數(shù)據(jù)，通過鏈?zhǔn)絽^(qū)塊結(jié)構(gòu)實現(xiàn)數(shù)據(jù)正向記錄和逆向追溯。世界狀態(tài)包括狀態(tài)數(shù)據(jù)庫、區(qū)塊索引數(shù)據(jù)庫歷史索引數(shù)據(jù)庫，追溯數(shù)據(jù)以鍵值對的形式存儲在狀態(tài)數(shù)據(jù)庫，將返回的交易哈希、區(qū)塊高度、區(qū)塊哈希等存放在索引數(shù)據(jù)庫，對同一鍵值數(shù)據(jù)執(zhí)行的更新操作存放歷史索引數(shù)據(jù)庫。索引數(shù)據(jù)庫通過索引指向區(qū)塊鏈文件系統(tǒng)區(qū)塊位置，確保鏈上僅最新數(shù)據(jù)可查。狀態(tài)數(shù)據(jù)庫為鏈外保存的CouchDB數(shù)據(jù)庫，存放帳本中所有鍵值對的當(dāng)前值，通過智能合約執(zhí)行交易時，可通過世界狀態(tài)索引查詢鍵值的最新狀態(tài)，無需全鏈遍歷查詢，提高數(shù)據(jù)查詢效率。

3.2 測試環(huán)境

測試環(huán)境基于Hyperledger Fabric搭建，使用虛擬機的配置為：32 G內(nèi)存、16核處理器、100 G硬盤，64萬條追溯試驗數(shù)據(jù)來源于山西忻州雜糧出口平臺。其中，生產(chǎn)鏈、加工鏈、倉儲鏈、物流鏈、分銷鏈，每條鏈通過4個節(jié)點存儲追溯數(shù)據(jù)，追溯鏈包括監(jiān)管節(jié)點、消費者節(jié)點以及企業(yè)節(jié)點；所有節(jié)點采用外部狀態(tài)數(shù)據(jù)庫CouchDB，通過索引檢索狀態(tài)數(shù)據(jù)庫代替遍歷查詢；所有測試用例以及測試結(jié)果通過區(qū)塊鏈基準(zhǔn)測試工具Hyperledger Caliper生成，試驗環(huán)境具體的配置如表4。

表4 區(qū)塊鏈配置

3.3 應(yīng)用案例分析

通過對雜糧產(chǎn)品供應(yīng)鏈實地調(diào)研，基于Hyperledger Fabric開發(fā)了區(qū)塊鏈追溯系統(tǒng)正向記錄“產(chǎn)、加、儲、運、銷”全過程追溯數(shù)據(jù)。系統(tǒng)應(yīng)用在山西忻州雜糧企業(yè)，該企業(yè)涉及雜糧供應(yīng)鏈所有追溯環(huán)節(jié)，但是數(shù)據(jù)存儲在追溯企業(yè)中心數(shù)據(jù)庫，數(shù)據(jù)易被篡改且難以即時動態(tài)監(jiān)管，因此，應(yīng)用本文開發(fā)的區(qū)塊鏈多鏈雜糧可監(jiān)管追溯系統(tǒng)進(jìn)行優(yōu)化。通過本文設(shè)計的區(qū)塊鏈追溯系統(tǒng)，雜糧產(chǎn)品貼有追溯二維碼，如圖4a所示。消費者通過手持設(shè)備掃描二維碼獲取產(chǎn)品全供應(yīng)鏈追溯數(shù)據(jù)，如圖4b所示。雜糧追溯數(shù)據(jù)存儲在區(qū)塊鏈賬本，追溯系統(tǒng)向用戶展示區(qū)塊鏈數(shù)據(jù)存儲信息，包括區(qū)塊鏈服務(wù)平臺訪問地址、鏈上追溯碼、區(qū)塊高度等，如圖4c所示。雜糧區(qū)塊鏈追溯服務(wù)平臺記錄交易數(shù)量3萬多筆，生成區(qū)塊1萬多，監(jiān)管部門通過鏈上不可篡改交易記錄監(jiān)管雜糧供應(yīng)鏈追溯企業(yè)數(shù)據(jù)交易行為，如圖4d所示。監(jiān)管部門通過服務(wù)平臺可監(jiān)管查看追溯數(shù)據(jù)，通過區(qū)塊高度查詢產(chǎn)品鏈上存證信息，如圖 4e所示。目前，忻州建成6塊、共計746.67 hm2的有機旱作區(qū)塊鏈追溯基地，打通農(nóng)業(yè)與區(qū)塊鏈發(fā)展壁壘，通過優(yōu)化追溯系統(tǒng)，雜糧產(chǎn)品質(zhì)量有所保證、銷量有所提升，其中玉米雜糧比較未采用區(qū)塊鏈追溯系統(tǒng)前售價提升28.8個百分點。

4 性能與安全性分析

4.1 區(qū)塊鏈追溯網(wǎng)絡(luò)性能分析

雜糧追溯網(wǎng)絡(luò)采用Raft共識機制，由全部追溯節(jié)點選取主節(jié)點進(jìn)行記賬、驗證、廣播，在主節(jié)點故障失聯(lián)后全網(wǎng)節(jié)點重新選取主節(jié)點，期間舊主節(jié)點操作回滾撤銷，增加網(wǎng)絡(luò)故障節(jié)點的容錯效率。圖5a展示將Hyperledger Fabric出塊時間設(shè)置為0.25～2.00 s時，追溯數(shù)據(jù)共識記賬時延。根據(jù)測試結(jié)果，數(shù)據(jù)記賬完成時延與出塊時間增長正相關(guān)，在出塊時間為0.50 s時數(shù)據(jù)共識記賬最大時延為0.82 s，平均時延為0.41 s，能夠滿足雜糧區(qū)塊鏈追溯的應(yīng)用需求。因此，本文設(shè)置出塊時間為0.50 s。

測試區(qū)塊鏈多鏈追溯網(wǎng)絡(luò)數(shù)據(jù)寫入與查詢吞吐量，如圖5b所示，在發(fā)送速率為25～150間網(wǎng)絡(luò)寫入吞吐量呈線性增加，在達(dá)到135后趨于平緩。在發(fā)送速率25～200間網(wǎng)絡(luò)讀取吞吐量呈線性增加，在達(dá)到200后趨于平緩。因此，本文建立的區(qū)塊鏈雜糧多鏈追溯網(wǎng)絡(luò)具有較高的數(shù)據(jù)寫入與查詢的吞吐量。

4.2 系統(tǒng)效率分析

驗證本文提出的區(qū)塊鏈雜糧多鏈追溯系統(tǒng)中追溯數(shù)據(jù)的寫入與查詢效率，為保證實驗數(shù)據(jù)的真實可靠性，所有測試結(jié)果均采用60次執(zhí)行結(jié)果的平均值。如圖6a所示，數(shù)據(jù)平均寫入時間為0.696 s；如圖6b所示，公開追溯數(shù)據(jù)平均查詢時間為0.415 s，敏感追溯數(shù)據(jù)平均查詢時間為0.871 s。從測試結(jié)果可以看出，面向監(jiān)管建立的雜糧供應(yīng)鏈多鏈追溯模型具有較好的數(shù)據(jù)查詢與寫入效率，能夠滿足雜糧銷售過程中的數(shù)據(jù)上鏈與查詢需求。

4.3 安全性分析

供應(yīng)鏈企業(yè)基于監(jiān)管授權(quán)組網(wǎng)建鏈以及操作鏈上賬本，因此，區(qū)塊鏈鏈上監(jiān)管安全性源于授權(quán)密文的安全性。以下對授權(quán)密文進(jìn)行相關(guān)性與擴散性測試，相關(guān)性測試是在授權(quán)密鑰不變的基礎(chǔ)上，通過修改授權(quán)內(nèi)容引起的密文變化情況，如圖7a所示，相關(guān)性測試中授權(quán)密文平均變化率為82.39%。擴散性測試是在授權(quán)內(nèi)容不變的基礎(chǔ)上，通過修改授權(quán)密鑰引起的密文變化情況，如圖7b所示，擴散性測試中授權(quán)密文平均變化率為82.53%。因此，該研究提出的授權(quán)算法具備較高的混淆性，能夠保證雜糧溯源網(wǎng)絡(luò)數(shù)據(jù)傳輸過程密文與密鑰的安全性。

5 結(jié) 論

本文設(shè)計并實現(xiàn)面向監(jiān)管的雜糧區(qū)塊鏈多鏈追溯模型，為規(guī)模化雜糧供應(yīng)鏈追溯提供借鑒與參考，并得出以下結(jié)論：

1）通過聯(lián)盟鏈Hyperledger Fabric通道技術(shù)構(gòu)建雜糧多鏈追溯系統(tǒng)，通過多條企業(yè)鏈并行操作提升網(wǎng)絡(luò)運行效率，擴展數(shù)據(jù)存儲容量，通過通道天然隔離性實現(xiàn)數(shù)據(jù)隱私保護(hù)；提出基于監(jiān)管部門授權(quán)的企業(yè)組網(wǎng)建鏈機制，監(jiān)管部門根據(jù)鏈上數(shù)據(jù)賬本實時審計結(jié)果管控追溯節(jié)點賬本操作權(quán)限；設(shè)計追溯數(shù)據(jù)上鏈合約，實現(xiàn)數(shù)據(jù)的鏈前監(jiān)管與鏈上管控；并設(shè)計數(shù)據(jù)查詢智能合約實現(xiàn)面向消費溯源與監(jiān)管溯源的差異化數(shù)據(jù)查詢。本研究提出的監(jiān)管授權(quán)方案，相關(guān)性測試中密文平均變化率為82.39%，擴散性測試中密文平均變化率為82.53%；經(jīng)區(qū)塊鏈追溯網(wǎng)絡(luò)吞吐量測試，消費者查詢公開追溯數(shù)據(jù)平均時間為0.415 s，監(jiān)管調(diào)用跨鏈接口查詢企業(yè)追溯數(shù)據(jù)平均時間為0.871 s。

2）構(gòu)建并實現(xiàn)的雜糧供應(yīng)鏈追溯系統(tǒng)能夠解決追溯鏈上多企業(yè)間數(shù)據(jù)隱私保護(hù)、數(shù)據(jù)差異化共享以及追溯數(shù)據(jù)的穿透式監(jiān)管、追溯節(jié)點的實時管控問題，實現(xiàn)了雜糧追溯大數(shù)據(jù)價值的深度挖掘、追溯網(wǎng)絡(luò)的安全可持續(xù)發(fā)展。

[1] 李順國，劉斐，劉猛，等. 中國谷子產(chǎn)業(yè)和種業(yè)發(fā)展現(xiàn)狀與未來展望[J]. 中國農(nóng)業(yè)科學(xué)，2021，54(3)：459-470.

Li Shunguo, Liu Fei, Liu Meng, et al. Current status and future prospective of foxtail millet production and seed industry in china[J]. Scientia Agricultural Sinica, 2021, 54(3): 459-470. (in Chinese with English abstract)

[2] Thakali A, Macrae J D. A review of chemical and microbial contamination in food: What are the threats to a circular food system?[J]. Environmental Research, 2021, 194: 110635.

[3] 王新平，柴尚森. 基于MCDM的食用農(nóng)產(chǎn)品追溯體系效能影響因素分析[J/OL]. 食品科學(xué)，[2021-05-28]，http: //kns. cnki. net/kcms/detail/11. 2206. ts. 20210305. 1635. 002. html.

Wang Xinping, Chai Shangsen. Analysis of factors influencing the effectiveness of edible agricultural products traceability system based on MCDM[J/OL]. Food Science, [2021-05-28], http: //kns. cnki. net/kcms/detail/11. 2206. ts. 20210305. 1635. 002. html. (in Chinese with English abstract)

[4] 涂海寧，魏俊文，劉建勝，等. 條碼圖像模式識別的追溯系統(tǒng)設(shè)計與實現(xiàn)[J]. 機械設(shè)計與制造，2021(7)：241-245.

Tu Haining, Wei Junwen, Liu Jiansheng, et al. Design and implementation of traceability system for bar code image pattern recognition[J]. Machinery Design & Manufacture, 2021(7): 241-245. (in Chinese with English abstract)

[5] 鄭立華，冀榮華，王敏娟，等. 農(nóng)產(chǎn)品追溯統(tǒng)一編碼方案設(shè)計與應(yīng)用[J]. 農(nóng)業(yè)機械學(xué)報，2019，50(S1)：385-392.

Zheng Lihua, Ji Ronghua, Wang Minjuan, et al. Design and application of traceable unified coding scheme for agricultural products[J]. Transactions of the Chinese Society for Agricultural Machinery, 2019, 50(S1): 385-392. (in Chinese with English abstract)

[6] 王祖良，郭建新，張婷，等. 農(nóng)產(chǎn)品質(zhì)量追溯RFID標(biāo)簽批量識別[J]. 農(nóng)業(yè)工程學(xué)報，2020，36(10)：150-157.

Wang Zuliang, Guo Jianxin, Zhang Ting, et al. RFID batch-tag identification in quality traceability of agricultural products[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2020, 36(10): 150-157. (in Chinese with English abstract)

[7] 錢建平，吳文斌，楊鵬. 新一代信息技術(shù)對農(nóng)產(chǎn)品追溯系統(tǒng)智能化影響的綜述[J]. 農(nóng)業(yè)工程學(xué)報，2020，36(5)：182-191.

Qian Jianping, Wu Wenbin, Yang Peng. Review on agricultural products smart traceability system affected by new generation information technology[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2020, 36(5): 182-191. (in Chinese with English abstract)

[8] 孫傳恒，于華竟，徐大明，等. 農(nóng)產(chǎn)品供應(yīng)鏈區(qū)塊鏈追溯技術(shù)研究進(jìn)展與展望[J]. 農(nóng)業(yè)機械學(xué)報，2021，52(1)：1-13.

Sun Chuanheng, Yu Huajing, Xu Daming, et al. Review and prospect of agri-products supply chain traceability based on blockchain technology[J]. Transactions of the Chinese Society for Agricultural Machinery, 2021, 52(1): 1-13. (in Chinese with English abstract)

[9] Yang X T, Li M Q, Yu H J, et al. A trusted blockchain-based traceability system for fruit and vegetable agricultural products[J]. IEEE Access, 2021, 9: 36282-36293.

[10] 楊信廷，王明亭，徐大明，等. 基于區(qū)塊鏈的農(nóng)產(chǎn)品追溯系統(tǒng)信息存儲模型與查詢方法[J]. 農(nóng)業(yè)工程學(xué)報，2019，35(22)：323-330.

Yang Xinting, Wang Mingting, Xu Daming, et al. Data storage and query method of agricultural products traceability information based on blockchain[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2019, 35(22): 323-330. (in Chinese with English abstract)

[11] 于合龍，陳邦越，徐大明，等. 基于區(qū)塊鏈的水稻供應(yīng)鏈追溯信息保護(hù)模型研究[J]. 農(nóng)業(yè)機械學(xué)報，2020，51(8)：328-335.

Yu Helong, Chen Bangyue, Xu Daming, et al. Modeling of rice supply chain traceability information protection based on block chain[J]. Transactions of the Chinese Society for Agricultural Machinery, 2020, 51(8): 328-335. (in Chinese with English abstract)

[12] Yang Z L, Li X, He P. A decision algorithm for selecting the design scheme for blockchain-based agricultural product traceability system in q-rung orthopair fuzzy environment[J]. Journal of Cleaner Production, 2021, 290: 125191.

[13] 李宣，柳毅. 基于雙區(qū)塊鏈及物聯(lián)網(wǎng)技術(shù)的防偽追溯系統(tǒng)[J]. 計算機應(yīng)用研究，2020，37(11)：3401-3405， 3421.

Li Xuan, Liu Yi. Traceability system based on double blockchain and Internet of Things technology[J]. Application Research of Computers, 2020, 37(11): 3401-3405, 3421. (in Chinese with English abstract)

[14] Cao Y B, Cheng Y Y, Yang C Y, et al. Double chain blockchain based on improved virginia algorithm[J]. Journal of Physics: Conference Series, 2021, 1820(1): 012158.

[15] Ding Q Y, Gao S, Zhu J M, et al. Permissioned blockchain-based double-layer framework for product traceability system[J]. IEEE Access, 2019, 8: 6209-6225.

[16] 李夢琪，楊信廷，徐大明，等. 基于主從多鏈的水產(chǎn)品區(qū)塊鏈溯源信息管理系統(tǒng)設(shè)計與實現(xiàn)[J]. 漁業(yè)現(xiàn)代化，2021，48(3)：80-89.

Li Mengqi, Yang Xinting, Xu Daming, er al. Design and implementation of aquatic product blockchain traceability information management system based on master-slave multi-chain[J]. Fishery Modernization, 2021, 48(3): 80-89. (in Chinese with English abstract)

[17] 陳純. 聯(lián)盟區(qū)塊鏈關(guān)鍵技術(shù)與區(qū)塊鏈的監(jiān)管挑戰(zhàn)[J]. 中國工業(yè)和信息化，2020(11)：54-58.

[18] 洪學(xué)海，汪洋，廖方宇. 區(qū)塊鏈安全監(jiān)管技術(shù)研究綜述[J]. 中國科學(xué)基金，2020，4(1)：18-24.

Hong Xuehai, Wang Yang, Liao Fangyu. Review on the technology research of blockchain security supervision[J]. Bulletin of National Natural Science Foundation of China. 2020, 34(1): 18-24. (in Chinese with English abstract)

[19] 張燕，侯啟玲. “區(qū)塊鏈+農(nóng)業(yè)眾籌”：創(chuàng)新、風(fēng)險及其法治監(jiān)管[J]. 華中農(nóng)業(yè)大學(xué)學(xué)報：社會科學(xué)版，2021(4)：137-145，184-185.

Zhang Yan, Hou Qiling. “Blockchain + agricultural crowdfunding”: invovation, risk and legal supervision[J]. Journal of Huazhong Agricultural University: Social Sciences Edition, 2021(4): 137-145, 184-185. (in Chinese with English abstract)

[20] 王毅. 基于區(qū)塊鏈的農(nóng)產(chǎn)品質(zhì)量安全追溯體系分析[J]. 青海農(nóng)林科技，2020(3)：59-61， 80.

Wang Yi. Analysis on traceability system of agricultural products quality safety based on the block chain[J]. Science and Technology of Qinghai Agriculture and Forestry, 2020(3): 59-61, 80. (in Chinese with English abstract)

[21] 許繼平，王健，張新，等. 區(qū)塊鏈驅(qū)動的稻米供應(yīng)鏈信息監(jiān)管模型研究[J]. 農(nóng)業(yè)機械學(xué)報：2021，52(5)：202-211， 101.

Xu Jiping, Wang Jian, Zhang Xin, et al. Information supervision modeling of rice supply chain driven by blockchain[J]. Transactions of the Chinese Society for Agricultural Machinery, 2021, 52(5): 202-211, 101. (in Chinese with English abstract)

[22] 霍紅，詹帥. 集成供應(yīng)鏈視角下農(nóng)產(chǎn)品質(zhì)量安全全過程監(jiān)管體系構(gòu)建[J]. 中國科技論壇，2019(8)：105-113.

Huo Hong, Zhan Shuai. Construction of a whole-process supervision system for the quality and safety of agrifood from the perspective of integrated supply chain[J]. Forum on Science and Technology in China, 2019(8): 105-113. (in Chinese with English abstract)

[23] 吳曉彤，柳平增，王志鏵. 基于區(qū)塊鏈的農(nóng)產(chǎn)品溯源系統(tǒng)研究[J]. 計算機應(yīng)用與軟件，2021，38(5)：42-48.

Wu Xiaotong, Liu Pingzeng, Wang Zhihua. Traceability system of agricultural products based on blockchain[J]. Computer Applications and Software, 2021, 38(5): 42-48. (in Chinese with English abstract)

[24] 孫知信，張鑫，相峰，等. 區(qū)塊鏈存儲可擴展性研究進(jìn)展[J]. 軟件學(xué)報，2021，32(1)：1-20.

Sun Zhixin, Zhang Xin, Xiang Feng, et al. Survey of storage scalability on blockchain[J]. Journal of Software, 2021, 32(1): 1-20. (in Chinese with English abstract)

[25] 梁昊，劉思辰，張一諾，等. 面向農(nóng)產(chǎn)品交易流程的多鏈?zhǔn)絽^(qū)塊鏈應(yīng)用技術(shù)研究[J]. 智慧農(nóng)業(yè)，2019，1(4)：72-82.

Liang Hao, Liu Sichen, Zhang Yinuo, et al. Multi-blockchain application technology for agricultural products transaction[J]. Smart Agriculture, 2019, 1(4): 72-82. (in Chinese with English abstract)

[26] Sanjeev K D, Ruhul A, Satyanarayana V. Blockchain-based secured ipfs-enable event storage technique with authentication protocol in vanet[J]. IEEE/CAA Journal of Automatica Sinica, 2021, 8(12): 1913-1922.

[27] 任守綱，何自明，周正己，等. 基于CSBFT區(qū)塊鏈的農(nóng)作物全產(chǎn)業(yè)鏈信息追溯平臺設(shè)計[J]. 農(nóng)業(yè)工程學(xué)報，2020，36(3)：279-286.

Ren Shougang, He Ziming, Zhou Zhengji, et al. Design and implementation of information tracing platform for crop whole industry chain based on CSBFT-blockchain[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2020, 36(3): 279-286. (in Chinese with English abstract)

[28] 王小兵，楊瀟鈺，舒新峰，等. 面向MSVL的智能合約形式化驗證[J]. 軟件學(xué)報，2021，32(6)：1849-1866.

Wang Xiaobing, Yang Xiaoyu, Shu Xinfeng, et al. Formal verification of smart contract based on MSVL[J]. Journal of Software, 2021, 32(6): 1849-1866. (in Chinese with English abstract)

[29] 張志威，王國仁，徐建良，等. 區(qū)塊鏈的數(shù)據(jù)管理技術(shù)綜述[J]. 軟件學(xué)報，2020，31(9)：2903-2925.

Zhang Zhiwei, Wang Guoren, Xu Jianliang, et al. Survey on data management in blockchain systems[J]. Journal of Software, 2020, 31(9): 2903-2925. (in Chinese with English abstract)

[30] Berdik D, Otoum S, Schmidt N, et al. A survey on blockchain for information systems management and security[J]. Information Processing & Management, 2021, 58(1): 102397.

Design of the blockchain multi-chain traceability supervision model for coarse cereal supply chain

Yu Huajing, Xu Daming, Luo Na, Xing Bin, Sun Chuanheng※

(1.,100097,; 2.,100097,)

Acoarse cereal is one of the most important crops rich in nutrients. However, the frequently-occurred issues on food quality and safety have seriously endangered the trust in recent years, for example, the abuse of chemicals, heavy metals exceeding the standard, and harmful germs. Therefore, a traceability system is an urgent need for coarse cereals, in order to bridge the trust between farms and consumers, where the source of agricultural coarse foods can be rapidly traced through multi-party efficient collaboration. Most current traceability systems focus on recording data using bar or QR codes, radio frequency identification, and wireless sensor networks. But the tracing information is broken easily between upstream and downstream, leading to the blur transferred data, particularly on the long and scattered supply-chain of agricultural foods, including cultivation, processing, logistics, storage, and sales. Furthermore, the traditional traceability system cannot efficiently supervise the data records of each company, when quality and safety issues occur. As such, it is impossible to recall the problematic foods in time, much less to accurately locate the responsible party. Fortunately, blockchain technology can be utilized to integrate the distributed architecture, storage, and verification in the block, peer-to-peer network protocols, encryption, consensus mechanisms, identity authentication, and smart contracts. The data disclosure can also be used to enhance trust with fewer intermediate links. Therefore, this study aims to deal with the long supply-chain of coarse cereals, many trace entities, the differentiate share of data ledger, and the real-time monitoring of on-chain data. A novel traceability system was designed to implement the supervisable food products using the multi-chain architecture of blockchain and supply chain in coarse cereal. The forward and reverse traceability data was also collected ranging from the planting, processing, warehousing, transportation, and sales, using cameras, sensors, Beidou positioning devices, and IoT devices in real-time. Moreover, a systematic investigation was made on the business process and supervision characteristics of the supply-chain, as well as the full life cycle of coarse cereal. In addition, a new architecture of supervision-oriented multi-chain data storage was proposed for the actual production of coarse cereal. Specifically, the off-chain CouchDB state database was selected to store the key-value traceability data, particularly on the key-index instead of traversal query to improve the query efficiency. More importantly, a network access mechanism was designed using the regulatory authority, further to realize the collaborative supervision on- and off-chain through smart contracts. Correspondingly, blockchain technology was used to ensure the deep connection of various production factor resources with the real economy of coarse cereal. Anyway, the supply chain of coarse cereal integrated with the blockchain was utilized to strengthen the multi-party collaboration through mutual identity authentication, especially on data expansion, data sharing, and supervision. In terms of security, the average change rate of ciphertext in the enterprise network authorization diffusivity test was 82.53%, the average change rate of ciphertext in the enterprise network authorization correlation test was 82.39%, indicating higher security and less confusion. In terms of efficiency, the average time for consumers to query public traceability data was 0.415 s, and the average time for regulators to call cross-link port to query enterprise sensitive traceability data was 0.871 s. Furthermore, an actual traceability system was implemented to verify the model using Hyperledger Fabric for data privacy protection, data differentiation sharing, and the penetration supervision of traceability data, together with the real-time management, and control of traceability nodes. The in-depth mining was realized for the value of coarse cereal traceability big data, and sustainable development of the traced network. As such, farmers can receive professional or industrial policy guidance, whereas, companies can obtain the market trends and price conditions in real-time, and regulatory authorities can accurately control traceability data and transaction behavior. The finding can greatly improve the quality and safety of coarse cereal, production efficiency, and economic benefits.

agricultural products; traceability; blockchain; multi-chain; data regulatory; Hyperledger Fabric

于華竟，徐大明，羅娜，等. 雜糧供應(yīng)鏈區(qū)塊鏈多鏈追溯監(jiān)管模型設(shè)計[J]. 農(nóng)業(yè)工程學(xué)報，2021，37(20)：323-332.doi：10.11975/j.issn.1002-6819.2021.20.036 http://www.tcsae.org

Yu Huajing, Xu Daming, Luo Na, et al. Design of the blockchain multi-chain traceability supervision model for coarse cereal supply chain[J]. Transactions of the Chinese Society of Agricultural Engineering (Transactions of the CSAE), 2021, 37(20): 323-332. (in Chinese with English abstract) doi：10.11975/j.issn.1002-6819.2021.20.036 http://www.tcsae.org

2021-06-02

2021-07-07

國家自然科學(xué)基金面上項目（31871525）

于華竟，研究方向為區(qū)塊鏈追溯技術(shù)。Email：18852897525@163.com

孫傳恒，博士，研究員，研究方向為農(nóng)產(chǎn)品追溯技術(shù)。Email：sunch@nercita.org.cn

10.11975/j.issn.1002-6819.2021.20.036

TP309.2；TS201.6

A

1002-6819(2021)-20-0323-10