Control of Mechanical System with Knowledge Data Base

知识数据库机械系统控制

• 批准号: 62420028
• 负责人: MIURA Hirofumi
• 金额: $ 12.29万
• 依托单位: The University of Tokyo
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for General Scientific Research (A)
• 财政年份: 1987
• 资助国家: 日本
• 起止时间: 1987 至 1989
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/grant/KAKENHI-PROJECT-62420028/
• 关键词: Intelligent Robot; Intelligent Traveling; Learning Ability; Knowledge Data Base; 知識デ-タベ-ス; 知識データベース

In this research, two kinds of mechanical system have been treated. The results are described in the following.(1) Intelligent Mobile Mechanical System-The quadruped robots were constructed. They walked autonomously. In walking motion, there are so many parameters and the criteria function for deciding a good walking pattern is not known. We investigated and analyzed the walking motion of actual animals ( dogs were mainly investigated ) in detail. It was found that energy consumption is one of very important parameters for walking motion. Gait ( which legs swing in which order ), duty factor ( ratio of the time the leg is contacting the floor over one cycle of walk ) and other parameters are decided by the criteria to minimize the energy consumption. This was concluded by comparison of investigating results of actual animals and results of dynamic analysis of the model. By control algorithm using this knowledge, the quadrupeds walked very smoothly in stable state.(2) Software for Intelligent Robot-We succeeded in constructing the robot having problem solving ability and learning ability. Problem solving means planning of action and execution of it. Learning means obtaining of necessary knowledges for problem solving. Robots which work in the real world have two problems. They are uncertainty of the result of action and lack of knowledge about working objects and environment. To conquer these problem, ability of problem solving and ability of learning requested. In this research, these two abilities were studied together. The results have been highly evaluated internationally.

在这项研究中，对两种机械系统进行了处理。研究结果如下：(1)智能移动机械系统——构建了四足机器人。他们自主行走。在步行运动中，有很多参数，并且决定良好步行模式的标准函数尚不清楚。我们对实际动物（主要是狗）的行走动作进行了详细的调查和分析。研究发现能量消耗是步行运动非常重要的参数之一。步态（哪条腿按什么顺序摆动）、占空因数（在一个步行周期中腿接触地板的时间比率）和其他参数由最小化能量消耗的标准决定。这是通过比较实际动物的研究结果和模型的动态分析结果得出的结论。通过利用这些知识的控制算法，四足动物在稳定状态下非常平稳地行走。(2)智能机器人软件——我们成功地构建了具有解决问题能力和学习能力的机器人。解决问题意味着行动的计划和执行。学习意味着获得解决问题所必需的知识。在现实世界中工作的机器人有两个问题。它们是行动结果的不确定性以及对工作对象和环境缺乏了解。要攻克这些问题，需要解决问题的能力和学习的能力。在这项研究中，这两种能力被一起研究。其成果得到了国际上的高度评价。

项目成果

木村浩,下山勲,三浦宏文: 第5回日本ロボット学会学術講演会論文集. 371-372 (1987)

Hiroshi Kimura、Isao Shimoyama、Hirofumi Miura：日本机器人学会第五届学术会议论文集 371-372（1987）。

H.Kemura, I.Shimoyama, H.Miura: "Criteria for Dynamic Walk of the Quadruped" Proceedings of THE INTERNATIONAL SYMPOSIUM AND EXPOSITION ON ROBOTS. 595-600 (1988)

H.Kemura、I.Shimoyama、H.Miura：“四足动物动态行走的标准”国际机器人研讨会暨博览会论文集。

J. Miura, I. Shimoyama and H. Miura: "Automatic Task Planning of Robots" Proc. of 5th Meeting of the Robotics Society of Japan 55-58, 1987.

J. Miura、I. Shimoyama 和 H. Miura：“机器人的自动任务规划”Proc。

三浦純,下山勲,三浦宏文: "ロボットの自動作業計画生成-実行システム構成のための枠組" 第5回日本ロボット学会学術講演会論文集. 55-58 (1987)

Jun Miura、Isao Shimoyama、Hirofumi Miura：“机器人自动工作计划生成 - 执行系统配置框架”日本机器人学会第五届学术会议论文集 55-58（1987）。

三浦純,下山勲,三浦宏文: "知能ロボットにおける問題解決と学習のモデル" 第6回日本ロボット学会学術講演会論文集. 183-186 (1988)

Jun Miura、Isao Shimoyama、Hirofumi Miura：“智能机器人的问题解决和学习模型”日本机器人学会第六届学术会议论文集183-186（1988）。