基于云模型的地面智能机器人自主性评价方法

Autonomy Evaluation Method of Ground Intelligent Robot Based on Cloud Model

为解决地面智能机器人自主性评价过程中评价体系单一、未考虑评价过程及系统自身不确定性等问题,提出了一种基于云模型的自主性评价方法。该方法建立了评价云模型,以期望作为评价基本度量,以熵和超熵描述评价的不确定性。通过对不同属性特征的样本进行统计特征、区间数和评价向量运算,得到实数型、区间型和语言型三类评价云生成方法。利用加性合并计算,将多个评价云信息累加,得到具有稳定可信性的综合评价状态云。以实验室3台地面智能机器人为原型,设计系统自主性评估验证实例,完成了云化过程。实验验证了该方法能够有效结合定性与定量评价,可应用于地面智能机器人自主性评价和其他复杂智能系统的综合性能评价。

To solve the problems of the autonomy evaluation process for ground intelligent robot that is prevalently oversimplified and barely concerns the uncertainty of the evaluation process and the system, an autonomy evaluation method is presented based on a cloud model. An evaluation cloud model is established, expectation is proposed as the basic metric of the autonomy standard, and entropy and hyper-entropy are proposed to describe the uncertainty of the evaluation. The generation methods of three types of evaluation clouds including real number type, interval number type and language type are obtained by the calculation of statistical features, interval-number features and evaluation vector features for different attributes of samples. A synthesized evaluation cloud with stable reliability is generated through computing with multi characteristics of clouds using additive property calculation. The cloud process is completed through designing verification cases of the system autonomy evaluation using three ground intelligent robots in laboratory as archetypes. The results show that this method combines the qualitative and quantitative evaluation effectively, and it is feasible in the autonomy evaluation of ground intelligent robots and the comprehensive performance evaluation of other complex intelligent systems.