本文將新產品開發的計畫區分為多種研發(R&D)類別。每種研發類別代表著一種可以再設計/升級的特定產品。類別當中的每項專案執行品質(技術)都擁有多重的等級可供選擇。鑑此，本文提出一個多類別及其多等級的專案選擇性問題，該問題亦同時考量每期預算限制下的專案排程議題。本文的內容係由四個部分組成：(1)專案推移策略的選擇係專案排序時考量軟性因素做為專案排程的架構，(2)將消費者對產品的品牌印象引用為長期的平均獲利最大化之目標函數，(3)將週期性預算限制及模糊與隨機之價值期限轉換為一可計算之模式，(4)轉換目標函數轉換為一適當格式，以使參數更容易被估計，進而使最終結果被解釋為具管理意涵的內容。另外，我們針對本文所提出之每個模型，逐行進行54項模擬測試。測試結果發現模式的最佳解能夠在短時間內被找到；換言之，本文所提出之模型是具有可行性與應用性。;This research proposes a new approach to select projects and their quality-standards corresponding to successively providing new products in the market. The problem involves the situation in which each new product development (NPD) program can be differentiated as multiple categories. Each project has multiple choices of quality/technology standards. Based on this problem, the proposed approach consists of following four components: (1) selecting a project advancement strategy to serve as a scheduling framework for taking into account soft factors in scheduling process, (2) employing the brand-image score of consumers as the objective function for ultimately increasing long-run average profitability, (3) formulating a computable model in which periodical budget constraints are involved and fuzzy and stochastic value-based time limits are specified, and (4) transforming the objective function into an appropriate form in which the parameters can be estimated more easily and the objective value can be predicated as a clear managerial implication. Also, we would simulate 54 items to test the computable model. The testing results should can find the global optimal in a short time for most testing items. That is, and the proposed computable model has high feasibility and applicability.