Heuristic Algorithm for Scheduling of Make-to-Order Plant Factory with Considering Crops Size Variation during Cultivation

植物工廠為新式的農作物生產模式，透過電腦系統對生產環境因素，如溫度、濕度、光源、水分及營養物的完全控制，加速植物生長的效能並且提升產品品質。植物工廠相較於傳統農業營運成本高出許多，如何制定生產計劃提高資源的利用以增加植物工廠的營利成為經營的重要議題。考量農作物交易價格、環境限制、使用空間等，進行植物工廠的生產排程規劃，本研究針對訂單式生產之完全人工控制型植物工廠的排程問題進行求解，將此問題用混合整數規劃模型表示，目標為最大化總收益扣除總成本之後的總利潤，限制分別考慮植物工廠空間、栽培室環境、作物特性、訂單的交貨期限…等，進而求解最佳化模式。使用LINGO軟體求解此混合整數規劃模型可得到最佳解，但使用軟體求解的時間隨實驗規模變大呈現指數成長，故提出一個啟發式演算法(HOBPFS)，此演算法考量訂單價格與相關限制，以迴圈的方式對每筆訂單進行生產排程安排，並且在合理的時間內得到可行解，HOBPFS主要包含三個部分:訂單分析與栽培室環境分配、訂單生產排程規劃、可行解的改善。使用此演算法在實證研究上，實驗中設定四種實驗因素進行討論，分別為(1)訂單數量、(2)農作物種類、(3)栽培室數目與(4)是否考慮植物體積變化，且由數值分析能看出，農作物種類與栽培室數目具有相關性，而考慮植物體積變化之情況下確實提高收益符合預期，且使用此演算法得到的可行解的效果能達到最佳解的90%以上。Plant factory is an environmental controlled facility which can sustain the stable crop cultivation with fast production and better quality by controlling temperature, humidity, lighting, nutrient supply and other cultivating factors. In this research, focus on production scheduling problem of ordering based (Make-to-Order) plant factory and consider size change as time past of crops. The scheduling problem was formulated as a mixed integer programming (MIP) problem. The objective function is to seek the maximum revenue of the plant factory by considering several practical operating conditions such as cultivation room space, environmental constraints, crop, due day of order. The operating conditions are formulated as constraints, and the MIP problem was solved by LINGO programming model to obtain the optimal solution. The computation time of LINGO solver is exponentially increasing when the problem domain is larger. Therefore, we propose the heuristic algorithm (HOBPFS) to solve the large size problem instance, which result show that our heuristic algorithm can indeed get good feasible solutions