Abstract
由於科技的發展與進步，使得倉儲存系統的設計及運作方式產生革命性的 變化。 自動化倉儲系統 (automated warehousing system; AWS) 包括自 動化存取系統 (automated storage/retrieval system;AS/RS) 及旋轉料 架 (carousel storage system) ， 這些利用電腦系統來控制及驅動的系 統，被使用的情況是愈來愈普遍。一般而言，自動化倉儲存系統的投資成 本相當的高，動輒上億元，對於如此昂貴的系統，如何做好規劃及有效的 使用是一個重要的課程，對於舊有系統的標準產出量的設定，預估新系統 的產出量以供設計時的參考，均顯現出本研究的重要性。對於自動化存取 系統而言，以前的文獻研究，對於期望行程時間的估計大都假設存取機的 存取移動速度為均速，而現實上，存取機普遍存在著具有多段速度及加減 速度的事實。本論文即對此點提出考慮加減速度下的旅行時間估計式，隨 後並將之推廣到料架設計的問題上。 旋轉料架是一種特殊構造的料架， 非常適用於小物品的訂單揀選，然而，研究旋轉料架的訂單揀選問題的研 究論文卻是非常的少。本論文即針對此點，提出單一旋轉料架的訂單揀選 問題，及一個取物裝置服務兩個料架的訂單揀選問題。希望本論文的研究 成果能提供決策者訂定標準產出量及績效評估時之參考。 In the automated manufacturing system, the efficiency of robot Technological developments have revolutionized the design and operation of warehousing systems. Automated warehousing systems (AWS), including automated storage/retrieval system (AS/RS) and carousel storage system, driven and controlled by computer system, are becoming increasingly prevalent. The investment cost of AWS, in general, is very expensive. Efficient utilization of the system is required to obtain highly throughput performance. Either establishing throughput standards on current systems or estimating throughput performance for the first cut evaluations of AWS design configurations is desirable. Expected travel-time expressions have been developed in previous research efforts involving AS/RS on the basis of an assumption that the speed of the storage/retrieval (S/R) machine is constant. The speed profile of the S/R machine has several speed segments and acceleration/deceleration rates in practical application. The expected travel-time expressions of the S/R machine drive are presented in this thesis in compact forms via a consideration of the speed profile of the S/R machine. The rack configuration problem is also addressed. Carousel systems are the most applicable for small part order picking. Those systems have received little attention in previous research efforts. In the thesis, the order picking problems are investigated for the single carousel system and two carousels single server system. The results presented are hopefully in providing insights for the decision maker into performance evaluation and throughput standard establishment. It is also desirable to support information that aids in estimating throughput performance for first evaluations of warehousing system design configurations.