本文主要是根據Lam（1994）的文章中。提出一新的決策函數，並且推導出一貝氏抽樣計劃 (n，t，T，ε，P)。我們考慮一新的損失函數，它包含了檢測時間的成本，並根據此推演出一最適抽樣計劃（n/sub B/，t/sub B/，T/ sub B/，ε/sub B/，p/sub B/）的演算法，最後，列出最適解及和Lam（1994）的數值比較。 A new decision function is proposed and its Bayes sampling plan(n,t,T,ε,p) is also obtained. The proposed sampling plan includes th e Lam's (1994) plan as special case. A new loss is considered which includes the cost of unit time. An algorithm for determining an optimal sampling plan (n/sub B/,t/sub B/, T/sub B/,ε/sub B/, P/sub B/ ) is given. Some optimal solutions are tabulated. Some numerical comparisons among the optimal sampling rule in Lam (l994), the proposed rule and other alternative rule have also been made.