本研究針對汽車保險中，車種為自用小客車及要保人為個人的部分，進行險種投保的多目標決策模 式之研究。首先透過分析層級程序法(analytic hierarchy process, AHP)來求得專家對各險種影響因子之權 重，然後透過專家經驗得出各險種的推薦評分表，再配合要保人偏好來建立多目標規劃模式，最後透過 Lingo 軟體來得出一建議投保的險種組合。另外，本研究亦演進式地依據此模式來發展雛型系統，要保人 可以互動方式來輸入偏好，並獲得適當的汽車保險投保建議。最後，各種模擬實驗、管理及若則分析在 此系統上執行及測試，所得到之管理意涵(managerial implications)或指導原則可以作為未來要保人在類似 情境中的決策參考。

This study focuses on multi-objective insurance policy purchasing model of personal car. First, the weight of each influence factor given from several insurance experts and the recommendation score of each kind of insurance are obtained and computed by the analytic hierarchy process (AHP) and experts’ experience respectively. Then, based on the preference, budget and other constraints, multiple objectives, etc. of the insured, an intelligent purchasing model is formed and solved by a business software, Lingo. Finally, its solution can be suggested to the insured as a combinatorial insurance policy. Further, a prototype system is developed by the above model and evolutionary prototyping method in this study. The insured can interact with the web-based system for entering and editing all required data and then obtain a suggested insurance policy on line. Finally, various simulation experiments, managerial and what-if analysis, etc. are executed and tested on the system, and then their managerial implications and guidelines can be used as the references of decision making that the insured incur in similar situations of the future.