Game theory is a famous issue of expert decision making. The real Shapley value for cooperative games with fuzzy characteristic function has high performance in deal with cooperative games, which is an effective tool in deal with issues of game theory. The real Shapley value for cooperative games with fuzzy characteristic function is based on the level sets, which is the extent of fuzzy sets. However, the real Shapley value for cooperative games with fuzzy characteristic function cannot solve the evidential games problems. What is the real Shapley value for evidential games with fuzzy characteristic function is still an open problem. This paper proposes the real Shapley value for evidential games with characteristic function, which consists of level sets, the real evidential Shapley value, basic probability assignment function. The real Shapley value for evidential games with fuzzy characteristic function can solve the expert decision making issues under evidential environment, with the aid of basic probability assignment function. Meanwhile, the theorem of the proposed model has been discussed. Numerical examples has been applied to illustrate the effectiveness of the proposed model. The experimental results show that proposed model can obtain the real evidential Shapley value of a given evidential games and address the issues of expert decision making.

博弈论是专家决策的一个著名问题。具有模糊特征函数的合作博弈的真实Shapley值在处理合作博弈方面具有较高的性能，是处理博弈论问题的有效工具。具有模糊特征函数的合作博弈的真实 Shapley 值基于水平集，即模糊集的范围。然而，具有模糊特征函数的合作博弈的真实 Shapley 值并不能解决证据博弈问题。具有模糊特征函数的证据博弈的真实 Shapley 值是多少仍然是一个悬而未决的问题。本文提出了具有特征函数的证据博弈的真实Shapley值，它由水平集、真实证据Shapley值、基本概率分配函数组成。具有模糊特征函数的证据博弈的真实Shapley值可以借助基本概率分配函数解决证据环境下的专家决策问题。同时，讨论了所提出模型的定理。数值例子已被应用来说明所提出模型的有效性。实验结果表明，所提出的模型可以获取给定证据博弈的真实证据沙普利值，并解决专家决策问题。数值例子已被应用来说明所提出模型的有效性。实验结果表明，所提出的模型可以获得给定证据博弈的真实证据沙普利值，并解决专家决策问题。数值例子已被应用来说明所提出模型的有效性。实验结果表明，所提出的模型可以获得给定证据博弈的真实证据沙普利值，并解决专家决策问题。