On the basis of set operation on integrity bo - dy , taking advanta ge of the excellency provided abundant topolo - gical information by the ra dial edge data structure , applying the divide and conquer method for inha ncing the efficiency of the system , the boolean operation on integrity or unintegrity bodys is furnished 本文在正则形体集合运算算法的基础上，充分应用辐射边结构提供丰富拓扑信息的优点，运用了分而治之的编程思想，实现了形体间的正则、非正则布尔运算。
There are some innovation points in this paper : decomposes the enterprise competitive advantage , discusses the elements of enterprise competitive advantage from the aspect of practice ; secondly , makes the theory of competitive intelligence blend into the practice of competitive advantage , decomposes the traditional competitive intelligence work , and combines the work into the two kinds competitive advantage , also separately discusses competitive " intelligence ' s work mechanism 本文的主要创新之处在于：将竞争优势的理论进行分解，从实践的角度研究竞争优势形成的要素。其二，将竞争情报的思想融合至竞争优势的实践中，对传统的竞争情报工作进行分化，不是笼统的进行竞争情报的需求、收集和分析工作，而是与竞争优势的各个不同构成要素进行结合，分而治之的进行分析。
Different parts in the same seamless system perform different tasks , and the use of oo method becomes the key point in the system integration of netfeaf . by making full use of some strategies like " from top to down " , " divide and conquer " and " step by step " , with the help of oo method , netfeaf creates a new method to build the whole integrated system : using one class hierarchy map to fully stand for the structure of netfeaf , and different parts in netfeaf like cluster class library , fea class library , web fea class library and gravity wharf cad class library can also be further explained by their sub class hierarchies , then the complexity of building netfeaf has been decreased step by step , the whole building procedure becomes much easier , so oo method can really change complexity into simplicity Netfeaf系统依据自顶向下、分而治之以及层层推进的实现原则，采用了全新的系统构造思路，对系统的集成用整体类库组成图进行表述，对各组成部分通过子类库再进一步详细介绍，系统的整体实现被一层一层分解、细化，各组成部分基于统一的计算核心集成在同一个系统框架内，同时相互之间还保持一定的独立性，独成一体，正是因为深入利用了面向对象的分析与实现手段，整个复杂系统构成从概念上更有利于理解，集成更加紧凑，具体模块的实现也更加迅速高效，面向对象方法的变复杂为简单的策略艺术在系统的具体实现过程中得到了充分展示。
On the basis of the concepts of divide - and - conquer and rough - fine , dia build a set of algorithm models including two rough classifiers ( i . e . bdt and cmac ) and three fine classifiers ( i . e . clchfpp , cpn - wed , nne - ifm ) which integrate some techniques such as clustering , decision tree , neural networks , fuzzy logic and information fusion etc . and take advantage of them , so dia has a high accuracy and speed theoretically Dia系统中算法模型的设计基于“分而治之”和“由粗到精”的模式识别思想，具体构建了两种粗分器( bdt和cmac )及三种精分器( clc - fpp 、 cpn - wed和nne - ifm ) ，这些分类器的设计综合了聚类法、决策树、神经网络、模糊逻辑以及信息融合等技术的应用，并使其优势互补，因此在理论上具有识别精度高及速度快的优点。
