智能科学的现状与发展趋势

Present Status and Future Tendency of Intelligent Science

首先简要介绍了智能科学的发展背景,提出了智能科学研究中要明确智能科学和智能技术的研究范畴,对2者概念上的混淆易导致研究路线和方法上的错误。论述了智能科学发展的哲学基础,并指出了智能科学的发展路线。详细介绍了人工智能技术,分析了符号主义、行为主义、连接主义3种人工智能路线的优势与不足。提出了抽象智能和直觉智能的分层智能概念。直觉智能主要负责处理底层应激性简单输入信号;抽象智能主要负责处理基于符号的需要逻辑推理的具有内在含义的输入信息。针对不同层次的输入信息必须采用相适应的智能处理方法才能达到预期的效果。对复杂信息的输入必须采取综合多种人工智能路线的方法。重点论述了目前的研究热点计算智能的研究内容和方法。最后指出了因为基于传统计算机实现的智能技术发展的瓶颈,发展智能科学必须走人机结合的路线,通过实现生物电子体,充分发挥生物智能处理模糊信息的优势,最终实现智能工业化的发展前景。

Background of intelligent science is first introduced in this paper.The difference of concept is also analyzed between intelligent science and intelligent technology.Research scope of intelligent science and intelligent technology must be clearly defined,otherwise confusion of the two concepts will result in wrong research route and method.Philosophy basics of intelligent science are discussed and its development route is also derived.AI（artificial intelligence） is introduced particularly.The advantage and shortcoming are analyzed among the three AI method： characteristic,connectism and behaviorism.Abstract intelligence and direct intelligence based on hierarchy structure intelligence are brought out.Direct intelligence mainly deals with lower lever and simple inputted signal. Abstract intelligence mainly deals with meaningful and logical reasoning information based on character.Different method would be adopted for different level information.For complicated information,multi-method with multi-route of AI would be integrated.Study scope and methods of the study hotpot,computing intelligence,are discussed in great detail.At last it is pointed out that human-computer combination route is the best way to develop intelligent science because there is a big bottleneck in the traditional intelligent technology based on computer method.Intelligent industry will be realized by bringing the advantage of fuzzy information processing ability of biological intelligence into full play through the bio-electronic system.