應用區塊鏈於灌溉水質紀錄與污染溯源──以桃園大圳二、三、四支線為例

Abstract

為了解決傳統資料庫有可能被竄改因此無法取得使用者信任的問題，本文提出了一個區塊鏈的架構，用區塊鏈儲存灌溉水質監測網的資料。區塊鏈有不可竄改、資料透明與交易可溯源的特性，因此非常適合用於儲存灌溉水質的監測資料。本研究將回顧區塊鏈的基礎原理，並提出一個完整的架構，使得原本用於加密貨幣的區塊鏈模型能被應用於灌溉水質監測網。區塊鏈是建立在對等式網路上的分散式帳本，每一個參與區塊鏈的電腦稱為「節點」，這些節點間彼此相連，並各自儲存有一份完整的資料，因此被稱作是去中心化的網路。區塊鏈會將一段時間內的交易包裝成一個區塊，對區塊做雜湊函數，並將雜湊的結果儲存至下一個區塊，區塊與區塊間彼此被鍊結，因此被稱作為區塊鏈。因為不可竄改的特性，除了加密貨幣外，區塊鏈已經被運用於許多領域。 研究區域內的測站包含基本測站與重金屬測站，基本測站監測水溫、電導度與pH，重金屬測站除了基本測站的監測項目外，另有監測鎘、銅、鉛、鎳、鋅、鉻六項重金屬的濃度。本研究使用GCOIN區塊鏈實作溯源系統，將加密貨幣模型中的「幣」對應到污染紀錄。每一個測站有一個對應的區塊鏈的地址，當測站監測到污染時，得到一枚幣，如果這個污染是來自其上游的測站，則將此枚幣送往其上游的測站。區塊鏈提供交易溯源的機制，透過查詢幣的交易歷史，可以知道測站發生污染的紀錄。上鏈時，可參考上鏈流程圖，根據上下游關係對測站排序後，逐測站操作。每筆交易都包含一個OP_RETURN的輸出，用字串的方式儲存監測數值。查詢時以樹狀圖呈現交易的歷程與監測值，所有資料皆取自區塊鏈，以區塊鏈確保資料的正確性。 本文以2018年4月的每一個整點時的監測值上鏈，並討論各種超標情形的組合，驗證本區塊鏈架構足以使用於水質監測網。對於區塊鏈常被質疑的可擴展性問題，本文亦有測試儲存空間的需求與每次上鏈所需時間。可擴展性測試的結果顯示，就灌溉水質監測網的規模，可以使用區塊鏈加強資料的安全性。 本文提出的方法有兩個侷限：1. 並未嚴謹地討論兩側站之間的污染傳遞，當緊鄰且有上下游關係的兩測站同時超標，即判定為同一污染事件，未來可以結合相關水力模式解決此問題；2. 區塊鏈僅記錄測站的監測值，若要真正找出汙染源，需要結合其他工具。

One of the problems that frequently occurred in the traditional database is that there is a high risk for it to be tampered and thus cannot win trust from users This thesis thus provides a solution by proposing a blockchain framework which uses blockchain technique to store the data of the irrigation water quality monitoring network. There are a lot of advantages of blockchain technique, including immutability, data transparency and transaction traceability. Those advantages make blockchain ideal for monitoring data on irrigation water quality. The earliest application of the blockchain is cryptocurrency. This study will review the basic principles of blockchain and propose a complete frameworkthat allows the blockchain model to be used for irrigation water quality monitoring networks. Blockchain is a decentralized ledger based on a peer-to-peer network. Each computer joining in a blockchain is called a "node". Those nodes are connected to one other and the data is stored in the ledgers completely. It is therefore called a decentralized network. The block chain will package the transactions over a period of time into a block, create a hash value on the block, and store the hashed results in the next block. The block is chained to next block, so it is called blockchain. Because of its immutable nature, blockchain technique has been applied in many areas already. The monitoring stations in the study area include regular stations and heavy metal stations. The regular stations monitor water temperature, conductivity and pH. On the other hand, in addition to those which are measured at regular stations, heavy metal stations also monitor the concentration of cadmium, copper, lead and nickel. This study uses the GCOIN blockchain implementation traceability system and maps the “coin” in the cryptocurrency model for the pollution record. Each station has a blockchain address. When the station monitors the pollution, a coin will be sent to the station.. If it is detected that the pollution event come from its upstream, the coin is sent to its upstream counterpart. The blockchain provides a mechanism for transaction traceability. By viewing the transaction history of the coins, users can know the record of pollution. Figure 16 provides a flow chart for uploading data onto the blockchain. After sorting the stations according to the upstream and downstream relationship, one can conduct the operation from those stations step by step, from downstream to upstream. Each transaction contains an OP_RETURN output, which stores the monitored values with a string. In the query, the history of the transaction is presented in a tree diagram with monitoring value. All the data are taken from the blockchain, in which the correctness of the data is ensured. This thesis takes the monitoring value of each hour point in April 2018 and discusses various combinations of exceeding standard cases to verify that the blockchain technique serves as a sufficient tool for the water quality monitoring network. For the scalability problem that are usually questioned regarding blockchain technique, this thesis also discusses the storage space demand and time required for each upload. The results show that the blockchain can be used to enhance the security of the data. The method proposed in this thesis has two limitations. 1. The pollution transmission between the two stations is not strictly discussed. When upstream stations and downstream station exceed standard simultaneously, it will be detected as the same pollution event. The hydraulic mode can be combined in the future to solve this problem; 2. The blockchain system only records the monitoring values of the station. It is necessary to combine other tools to further detect and define the actual source of pollution.