智能系统控制课件:enisc1.ppt
- 【下载声明】
1. 本站全部试题类文档,若标题没写含答案,则无答案;标题注明含答案的文档,主观题也可能无答案。请谨慎下单,一旦售出,不予退换。
2. 本站全部PPT文档均不含视频和音频,PPT中出现的音频或视频标识(或文字)仅表示流程,实际无音频或视频文件。请谨慎下单,一旦售出,不予退换。
3. 本页资料《智能系统控制课件:enisc1.ppt》由用户(罗嗣辉)主动上传,其收益全归该用户。163文库仅提供信息存储空间,仅对该用户上传内容的表现方式做保护处理,对上传内容本身不做任何修改或编辑。 若此文所含内容侵犯了您的版权或隐私,请立即通知163文库(点击联系客服),我们立即给予删除!
4. 请根据预览情况,自愿下载本文。本站不保证下载资源的准确性、安全性和完整性, 同时也不承担用户因使用这些下载资源对自己和他人造成任何形式的伤害或损失。
5. 本站所有资源如无特殊说明,都需要本地电脑安装OFFICE2007及以上版本和PDF阅读器,压缩文件请下载最新的WinRAR软件解压。
- 配套讲稿:
如PPT文件的首页显示word图标,表示该PPT已包含配套word讲稿。双击word图标可打开word文档。
- 特殊限制:
部分文档作品中含有的国旗、国徽等图片,仅作为作品整体效果示例展示,禁止商用。设计者仅对作品中独创性部分享有著作权。
- 关 键 词:
- 智能 系统 控制 课件 enisc1
- 资源描述:
-
1、Donghua Univ.Donghua Univ.Professor Tao GongCollege of Info S&T2014/09Book: Artificial Immune System Based on Normal Model and Its Applications, Tsinghua University Press, 2011Network Course: http:/ Representation, Search Ch1 Representation, Search and Reasoning of Knowledgeand Reasoning of Knowledg
2、eISCTao GongCollege of Information Science and Technology, DHU2CH1CH1 Representation, Search and Representation, Search and Reasoning of KnowledgeReasoning of KnowledgeConcepts and Designs of Expert SystemsSearch and Reasoning of KnowledgeKnowledge Representation of symbolicismThree Major Groups of
3、AIISC3Nowadays, the major groups of artificial intelligence (AI) include the following three ones:(1)Symbolicism, or called logicism, psychologism, computerism: its principles include the physical symbol system (i.e. symbol operation system) hypothesis and limited rationality principle.(2)Connection
4、ism, or called bionicsism, physiologism: its principles include neural networks, the connection mechanism among the neural networks, and learning algorithms.(3)Actionism, or called evolutionism, cyberneticsism: its principles include cybernetics and sensing-action control system.1 Three Major Groups
5、 of AICh1 Representation, Search & Reasoning of KnowledgeCollege of Information Science and Technology, DHUTao GongISC4 Symbolicism thinks that the AI originated from the mathematical logic.The mathematical logic developed quickly since the end of the 19th century, and it was used to describe the in
6、telligent behaviors since 1930s.After the invention of the computer, the computer was used to implement the logic deduction system. The heuristic program LT logic theorist was one of the typical fruits, and this program proved 38 mathematic theorems. This fact shows that the computer can be used to
7、investigate the thinking process of human beings and simulate their intelligent activity. Another example is the mechanical proving of geometry theorems by Prof. Wenjun Wu, and he won the top science award of China.Heuristic algorithmsExpert systemsKnowledge engineering1.1 Symbolicism1 Three Major G
8、roups of AICollege of Information Science and Technology, DHUTao GongISC5 Connectionism thinks the AI originated from the bionics, especially such as the research on the brain.The typical fruit of connectionism was the brain model (MP model), which was proposed by the physiologist McCulloch and the
9、mathematical logician Pitts in 1943. This model opened a new path of simulating the structure and functions of the brain with electric devices.MP modelperceptronHopfield neural networkBP algorithm1.2 Connectionism1 Three Major Groups of AICollege of Information Science and Technology, DHUTao GongISC
10、61.3 Actionism1 Three Major Groups of AICollege of Information Science and Technology, DHUTao Gong Actionism thinks that the AI originated from the cybernetics.The idea of cybernetics was popular from 1940s to 1950s, and this cybernetics was interesting for early researches (e.g. Wiener, Macollough
11、and Xuesen Qian) on artificial intelligence.From 1960s to 1970s, with the development of these control systems, the intelligent control and the intelligent robots were proposed. In 1980s, the intelligent control systems and the intelligent robots were designed and developed.The first typical fruit o
12、f actionism was the six-legged walking robot, which was designed by Brooks to simulate the insect behaviors with the control system based on the sensing-behavior pattern.The three major groups of AI will coexist and cooperate with each other for a long time by fusion and integration.ISC7 With on the
13、 symbols and logics, the traditional problem solving of AI is based on the knowledge representation and reasoning. Each knowledge-symbol-based intelligent system uses a searching procedure to find the solution for the problem. However, before the searching procedure, some knowledge representation ap
14、proaches should be utilized to represent the problem. These problem-representation approaches include the state space, the problem reduction, and the predicate formula etc. The right selection of suitable knowledge representation is important to increase the problem-solving efficiency of AI. The kno
15、wledge representation approaches use graphs, formulas, structures and statements etc.2 Knowledge RepresentationCh1 Representation, Search & Reasoning of KnowledgeCollege of Information Science and Technology, DHUTao GongISC8 The state space approach is a solution-space-based problem representation a
16、nd solving approach, which utilizes the states and the operators. When the knowledge is represented with the state space graph, the initial state is changed with one by one operator. The experimental series of the operators is built step by step, until the target state is attained. Because luxuriant
17、 nodes should be extended with the state space approach and this easily causes the combination explosion, this approach is only suitable for representing simple problems. The state is an ordered subset of the variables q0, q1, , qn , which describes the difference among the different objects of the
18、same class. Its vector is Q=q0,q1,qnT Here, the element qi (i=0,1,n) represents the state variable.2.1 State Space Approach2 Knowledge RepresentationCollege of Information Science and Technology, DHUTao GongISC9 The demo for solving the problem of the monkey and the banana is shown below. The right
19、text box shows the current state (W,x,Y,z). W denotes the horizontal position of the monkey; x has the value 1 when the monkey is on the box, otherwise x has the value 0; Y represents the horizontal position of the box; z has the value 1 when the monkey gets the banana, otherwise z has the value 0.C
20、ollege of Information Science and Technology, DHUTao Gong2.1 State Space Approach2 Knowledge RepresentationISC10 The problem reduction method is used from the object (the problem to be solved), through the backward inference and a series of transformations, to turn the initial problem into the set o
21、f sub-problems and the set of sub-sub-problems. These transformations end at the basic sub-problem set. The basic sub-problem can be solved directly, so that the initial problem can be solved. This problem-solving approach can be represented with the AND-OR graph. The state space approach is a speci
22、al example for the problem reduction method.2.2 Problem Reduction MethodCollege of Information Science and Technology, DHUTao Gong2 Knowledge RepresentationISC11College of Information Science and Technology, DHUTao Gong2.2 Problem Reduction Method2 Knowledge RepresentationISC12The predicate logic me
23、thod utilizes the appropriate formula of the predicate and one-order predicate calculation, in order to transform the problem solving into the problem proving. Afterwards, the resolution theorems and the resolution inversion are used to prove that a new sentence is derived from the known correct sen
24、tences, so this new sentence is also proved correct. The predicate logic is a kind of formal language, which can make the logical arguments of math symbolic.The predicate logic method is often used with other knowledge representation methods, and it is flexible and convenient to represent more compl
25、ex problems.College of Information Science and Technology, DHUTao Gong2.3 Predicate Logic Method2 Knowledge RepresentationISC13 语义网络是一种结构化表示方法,它由节点和语义网络是一种结构化表示方法,它由节点和弧线或链线组成。节点用于表示物体、概念和状态,弧线或链线组成。节点用于表示物体、概念和状态,弧线用于表示节点间的关系。弧线用于表示节点间的关系。 语义网络的解答是一个经过推理和匹配而得到语义网络的解答是一个经过推理和匹配而得到的具有明确结果的新的语义网络。的具有明
展开阅读全文