首页 / 专利库 / 软件 / 网络机器人 / Self-learning system

Self-learning system

阅读:910发布:2023-04-18

专利汇可以提供Self-learning system专利检索,专利查询,专利分析的服务。并且PURPOSE: To reduce the level of the noise to be added and to attain the safe operation of a system by controlling the noise applied to the output of an action network with use of the output of an evaluation network.
CONSTITUTION: When the sensor input is applied in an environment 33, the output of an action network 21 is calculated to the corresponding input pattern and then the output of an evaluation value learning network 22 is also calculated. A noise is obtained in response to the output of the network 22 and added to the network 21 and then stored in an input/output pattern short period storage part 23. The action output is outputted from an output part 28, and a robot moves in response to the action output. Then an evaluation signal 30 is applied via the environment 33 and based on the action result, and the network 22 learns the evaluation value. If the value of the signal 30 is decided satisfactory, an input/output pattern is learnt by the network 21 and a due action is repeated.
COPYRIGHT: (C)1991,JPO&Japio,下面是Self-learning system专利的具体信息内容。

高效检索全球专利

专利汇是专利免费检索,专利查询,专利分析-国家发明专利查询检索分析平台,是提供专利分析,专利查询,专利检索等数据服务功能的知识产权数据服务商。

我们的产品包含105个国家的1.26亿组数据,免费查、免费专利分析。

申请试用

分析报告

专利汇分析报告产品可以对行业情报数据进行梳理分析,涉及维度包括行业专利基本状况分析、地域分析、技术分析、发明人分析、申请人分析、专利权人分析、失效分析、核心专利分析、法律分析、研发重点分析、企业专利处境分析、技术处境分析、专利寿命分析、企业定位分析、引证分析等超过60个分析角度,系统通过AI智能系统对图表进行解读,只需1分钟,一键生成行业专利分析报告。

申请试用

QQ群二维码
意见反馈