物聯網技術於農業領域之監測系統開發與應用

Abstract

物聯網技術在近年開始被廣泛研究，並在各領域中已開始有各種應用實施案例，臺灣所提出的農業4.0方針，即是要以物聯網技術為基礎發展智慧科技農業。本研究首先針對物聯網核心技術進行研究與探討，提出一感測網高覆蓋路由演算法及低複雜度的定位演算法。高覆蓋率路由演算法可應用於高敏感區域之監測，定位驗算法可同時應用於集中式或分散式感測網路，依據應用端需求精度及響應時間，可以調度錨節點數量及演算複雜度。本研究分別以模擬及實作針對各類演算法進行驗證，結果顯示各演算法確實有效提升監測網路的運作時間，也可於室內場域達到定位成效，相關核心技術亦導入後續的實際應用中使用。 本研究運用機電整合技術，導入資通訊及物聯網相關技術應用於農業生態監測、害蟲監測及穀物儲藏管理監控。針對農業生態暨害蟲監測系統，本研究整合環境感測器、無線通訊晶片、微控制器等各式電子零組件，開發出適用於田間環境及害蟲監測之各式設備。所有監測資訊，能即時透過手機通信網路或區域無線通訊網路即時將資料回傳到資料庫系統內，使用者可透過網站平台即時監看資料。並可於網站平台上，進行歷史資料分析查閱、警報動作設定、取得蟲害預測資訊等進階功能操作。該系統在臺灣主要應用於東方果實蠅之自動化監測，可適用於露天栽培農業區、網室設施、有機農場等進行微氣候及果實蠅的棲群密度監測。透過即時且精確的監測系統，將有助於提升管理品質，並可透過後續資料分析，輔助管理者進行作業改善及提升作業效能。 歐美等地主要受到地中海果實蠅危害，也持續對地中海果實蠅實施誘引撲殺及人工監測，相關應用十分適合導入本研究所開發之農業生態暨害蟲監測系統使用。本研究首先針對地中海果實蠅設計專屬誘引通道，並導入已開發完成之監測設備與後臺系統，設計適用於地中海果實蠅之自動化監測系統。本研究與美國農業部太平洋研究中心的專家合作，試驗期間內多次於夏威夷大島及可愛島等地，進行地中海果實蠅監測試驗。試驗項目包括挑選適用自動化誘捕器之誘引劑、於野外環境進行自動化誘捕計數試驗、於網室內進行高密度誘捕試驗等。經多次試驗已找出適用於自動化誘捕計數裝置之蟲道尺寸，經野外試驗，驗證裝置可自動化計數地中 海果實蠅，且計數準確率達90%以上。且在大型網室內，以人工飼養的果實蠅進行高密度試驗，計數準確率可達70%以上。相關系統搭配各地的無線通訊系統後，即可進行長時間的自動化監測。 穀物倉儲因作業及管理方便之需求，國內目前大多將穀物存放於大型圓筒倉內。稻穀存放時必須控管穀物的溫度及濕度，才能確保稻穀碾製後的口感及風味。本研究導入物聯網感知技術及資通訊技術，建立智慧型穀物倉儲監控暨管理系統。針對監測需求，開發適用於筒倉的溫度線及濕度線、資料收集電路、管理平台及雲端分析平台等。本研究針對系統各式通訊功能完成性能測試及驗證，系統可依據未來使用者規模大小需求，彈性配置系統裝置及線路。監控系統可依據使用者需求，訂定穀物倉儲控溫區間、冷氣運作區間、異常通報模式等功能。透過導入智慧型穀倉監控與管理系統，未來將可提升穀倉系統管理之品質，除了能夠降低人工操作時造成的失誤損失，更能有效提升穀物存放時之保鮮率，並可降低穀物存放時之受損率，使產品更具有價值，進而能夠提高相關業者與農民之收益。 本研究所建立之各項物聯網監控系統，均已實際應用於各類農業領域，各系統可即時提供監測資訊供使用者作為生產管理修正依據，使農糧產品品質得以精進。針對長期的監測資料，後續更能透過大數據分析，找出有助於精進或改善各農業生產管理之方式，使農業生產作業管理更具效益，提升農產品品質及增加經濟效益。

In recent year, the researches for Internet of Things (IOTs) were increasing, and application of IOTs were used in a lot of area. The Agriculture 4.0 policy which is focused on intelligent agriculture based on the IOTs, was presented by Taiwan government. In the first, this research focused on the kernel technology of IOTs. The coverage preservation routing algorithm and low time complexity locating algorithm were presented. The upmost mission is to ensure that the network is fully functional providing reliable transmission of the sensed data without the risk of data loss. In this study, we propose a routing protocol to accommodate both energy-balance and coverage-preservation for sensor nodes in WSNs. The locating algorithm was able to apply in disturbed or centralized sensor networks. The number of anchor nodes and the time complexity can be adjusted based on the accuracy and response time of applications. The algorithms were evaluated by simulated and implementation, respectably. The results shows that the routing protocol can be improve working time for sensor network, and the localization protocol can be used in indoor localization application. Improving fruit farm profitability through integrated pest management (IPM) programs is always an important issue to modern agriculture systems. In order to enhance IPM programs against Bactrocera dorsalis, an automatic infield monitoring system is required to efficiently capture long-term and up-to-the-minute environmental fluctuations in a fruit farm. In this study, a remote agro-ecological monitoring system built upon IOTs has been developed to provide precision agriculture (PA) services with large-scale, long-distance, long-term, scalable, and real-time infield data collection capabilities. Historical data with spatial information is available through a web-based decision support program built upon a database. Pest population forecast results are also provided so that farmers and government officials would be able to accurately respond to infield variations. Compared with the previous version of the system, various useful functions have been added into the system, and its accuracy has been improved when measuring different parameters in the field. The system could provide a valuable framework for farmers and pest control officials to analyze the relations between population dynamics of the fruit fly and meteorological events. Based on the analysis, a better insect pest risk assessment and accurate decision-making strategy can be made as an aid to PA against B. dorsalis. Researchers could receive messages of the predicted data for prevent the outbreak of the fruit fly. Moreover, the pest management will be more efficient in the future. In this ex-site research, the pest population and attraction behavior of the Mediterranean fruit fly will be researched. The researchers work with Pacific Basin Agriculture Research Center, USDA. The automatic monitoring technology is constructed in this project, and we also establish the automatic monitoring system for the Mediterranean fruit fly in Hawaii, USA. The counting accuracy of automatic counting devices for the Ceratitits capitata (Mediterranean fruit fly) was presented. The researchers investigated the population and attraction behavior via video record and monitoring system for Mediterranean fruit fly. Moreover, the researchers done the tests in different population of the wild area (Coffee Farm, Hawaii) and also test in the cage of PBARC, USDA. In field tests, the counting accuracy is higher than 90%, and the counting accuracy is also higher than 70% with high density of medfly for the cage tests. Users are able to monitoring population of Mediterranean fruit fly by the automatic monitoring system, and the weather of the testing sites will be also monitored by the devices. Hence, the users could manage the farms by these monitoring data. Moreover, the population of Mediterranean fruit fly can be predict by historical monitoring data. Before the huge damage of Mediterranean fruit fly, the users could do some prevention measures. IOT were taken into the storage management for grain storages. Following with the designed grain storage monitoring system, the management system will be able to improve the fresh level and will decrease the damage level for the grain storage. This research has finished scheduled works about automatic management system for the grain silos, environmental monitoring system for grain silos, automatic monitoring platform for storage procedure, control strategy for temperature and humidity of silos, warning and alerting system for system fault. The monitoring database and central monitoring center for grain storage were also be establish. The users could monitor storage management system by web-based system, and users could setup different functions for the temperature control, chilling machine control, and model for alerting. The grain quality will be improvement, when the users start to use this management system for grain storage. The damage of grain storage will be also decrease, and the price of grain will be increase. The IOTs monitoring and control systems which were presented in this research were applied for agriculture solutions. The systems could offer real-time monitoring data for users, and the users are able to do different works according to the sensing and analysis data. The big data analysis and data mining will be applied for the long-term monitoring data, and the analysis results are able to improve the management methods and enhance the efficiency of the works. Moreover, the quality of agriculture products will be improvement, and the economic benefit will be also increased.