本文目的在探討談判者或決策者之間有討論時，談判者如何對特許契約所存在之不確定性因素進行風險衡量。本文以效用相依及偏好可分性(weakly separable preference)觀念，採動態規劃方法構建談判群體效用衡量模式，研擬疊代求解法(iterative algorithm)，並以範例分析方式說明模式之可用性。經由範例分析顯示，影響談判者之間效用獨立與否，取決於效用交互影響值(interactive utility value, IUV)之和是否為0；當效用交互影響值之和趨近於0時，談判者之間的效用相依程度愈低，獨立越強；相對地，當校用交互影響值之和趨近於1時，談判者之間的效用相依程度越高;當談判者之間的效用差異越小，模式易於收斂；若談判者之間的效用差異越大，討論次數會增加；討論次數增加時，談判者之間的效用交互影響值會遞減。而當在第l次討論即獲得求解時，此時談判群體投用值可由談判者之最初效用加總，此與談判者在獨立狀態下之衡量相同。

The purpose of this paper is to measure the risk of BOT projects when there is interactive relationship among the negotiators. Based on the utility theorem and the weak separability theorem, a dynamic programming of the risk measurement model is developed to simulate the utility dependent behavior among the negotiators. The results of numerical examples show that the inter, active relationship of negotiators increases when the sum of interactive utility value is near 1. Otherwise, the interactive relationship of negotiators becomes independent when the sum of interactive utility value is 0. The algorithm of the model can be converged when the difference value between negotiator s utility is small, and the discussion frequency increases when the difference value between negotiator s utility becomes large. It shows that the group utility model and iterative algorithm in this paper call be applied to analyze the inter, active behavior and risk measurement in BOT projects.