专门讨论了对智能系统的智能特性进行评价的问题。为此,提出了一种评价系统的评价模型及评价方法。力求排除其它影响因素。只关注系统的自识别等智能特性。采用广义算子模型对评价过程进行描述。并运用仿真和实验室实验的手段,以智能移动机器人对特定对象自识别智能特性的评价为例。给出实施测试时解决几个关键问题的方法和算法。结果表明:拥有人工智能的机器或系统的智能特性是可评价的,通过描述、测试和评价得出的定性评价结论可用于判断系统“智能特性”的有无,这是对系统进行智能水平的定量评价的前提。
This paper focuses on the evaluation of intelligent characteristics of intelligent control system.Therefore we bring forward evaluation models and methods of an evaluation system.We are trying to exclude other factors ,but concentrate on the intelligent characteristics such as the self-recognition and so on.We apply the model of generalized operator to describe the evaluation course,and provide solutions and algorithms for several critical problems in operation by using imitation and other lab measurements working on the evaluation of intelligent robot's intelligent characteristics in self-recognition of certain objects.The results show that it is possible to evaluate the intelligent characteristics for machines or systems with artificial intelligence.The qualitative conclusion through description,examination and evaluation could be applied in judging whether a system has intelligent characteristics or not,which is the precondition of quantitative evaluation for intelligence level of any given systems.