论文部分内容阅读
把外界刺激、知识库中的规则离散为基本元素并且和信念、期望及意图主体(BDIagent)的前提结论结合在一起,用可学习的函数关系表示了信念、期望及意图的相互转换机制。BDIagent通过生长和演化在规则库中留下痕迹,此痕迹对规则圈中的优先树进行修改,BDIagent在内部函数作用下生长变大,在外部函数作用下演化为友员主体,也可以蜕变为面向对象的类;演化生长为具有各种用途的、在一个整体前提下有机组合的各种功能单位。这种主体可方便地建造适应于许多智能领域的应用程序,使其成为面向自组织设计的基本概念与基本数据结构,以这种算法结构为基础能构造出自组织设计所需要的主体工作平台。
Discrete rules of external stimuli and knowledge base as the basic elements and combined with the preconditions of belief, expectation and BDIagent, and expressed the mutual conversion mechanism of beliefs, expectations and intentions through the function of learning. BDIagent leaves a trail in the rule base through its growth and evolution. This trail modifies the priority tree in the circle of rules. BDIagent grows larger under the function of internal function and evolves into a friend body under the function of external function. It can also be transformed into Object-oriented class; evolved into a variety of uses for the development, under the overall premise of a combination of various functional units. The main body can be easily adapted to many intelligent applications, making it a basic concept and basic data structure for self-organizing design. Based on this algorithm structure, the main work platform needed for self-organizing design can be constructed.