道路网智能选取的案例类比推理法

Intelligent Road-network Selection Using Cases Based Reasoning

从人类学习和认知角度,借鉴人工智能领域基于案例推理学习的成果,提出一种基于案例类比推理的道路网智能选取新方法。该方法将制图专家对某区域道路网的交互选取结果作为参考标准,对其进行结构化描述并构建和转化为案例库;计算机采用一定的简化算法和泛化算法对该案例库进行分析和学习,获取检索效率更高和适应样本能力更强的案例模型库;计算机在对相似道路网自动选取时,根据获取的案例模型库,采用基于案例类比推理的方法,分析获取相应的解决方案,进而完成道路网智能选取。与已有研究成果相比,本方法以案例及其泛化模型来模拟专家思维,以计算机对案例模型的类比学习来进行相似道路网自动选取,增强了道路网选取中的智能性。最后对本方法的科学性和适用性进行验证,并对试验结果作分析和评价,同时指出了存在的问题和进一步的研究方向。

A new approach of intelligent road-network selection using cases based reasoning(CBR)is put forward.In this approach,learning and cognition techniques of human being in artificial intelligence are used to establish,learn and reason the cases of cartographers.First,it takes a certain area's road-network selection result achieved from interactive selection of cartographic experts as reference templates,and transform the templates into selection cases after establishing the description structure of cases.Second,the cases are analyzed and reasoned with enhanced simplifying and generalizing methods so as to get more effective case model base.Finally,the computer finishes the similar road selection using CBR technique supported with the enhanced case model base.Compared with the past algorithms,the proposed approach uses enhanced road selection cases to simulate the thinking of human being,and CBR model to select similar road-work intelligently,which enhances the intelligence of traditional road selection methods.Examples and related analyzing and assessing results illustrate the scientificity and usability of the new approach.And further works to be improved are also suggested.