专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明专利

JP2013143009A Equipment state monitoring method and device using the same 审中-公开
标题翻译：设备状态监控方法和使用该设备的设备
公开(公告)号：JP2013143009A
公开(公告)日：2013-07-22
申请号：JP2012003027
申请日：2012-01-11
申请人： Hitachi Ltd , 株式会社日立製作所 , Wakayama Univ , 国立大学法人和歌山大学
发明人： SHIBUYA HISAE , MAEDA SHUNJI , WADA TOSHIKAZU , OZAKI SHINSAKU
IPC分类号： G05B23/02
摘要： PROBLEM TO BE SOLVED: To provide an equipment state monitoring method and a device using it that can detect abnormality with a high sensitivity while suppressing the calculation load at a low level, and can determine whether the sensor value at a given moment as well as the change of the sensor value between one time point and another are normal or abnormal, regardless of the way time changes.SOLUTION: A device using an equipment state monitoring method: inputs sensor signals output from a sensor mounted on equipment; acquires observation data by converting the input signals by a time direction scale; calculates abnormality at each time point by processing the observation data; and detects the abnormal period of the equipment on the basis of the calculated abnormality at each time point. The device using the equipment state monitoring method is designed to select a predetermined number of learning data close to the observation data among the pre-stored learning data, create a normal model by using the selected learning data, and calculate the abnormality of the observation data by using the observation data and the normal model.
摘要翻译：要解决的问题：提供一种设备状态监视方法和使用该设备状态监视方法的装置，其能够以低灵敏度抑制计算负荷而以高灵敏度检测异常，并且可以确定给定时刻的传感器值以及一个时间点和另一个时间点之间的传感器值的变化是正常的或异常的，而不管方式时间的变化。解决方案：使用设备状态监测方法的设备：输入从安装在设备上的传感器输出的传感器信号; 通过将输入信号转换为时间方向标度来获取观测数据; 通过处理观察数据来计算每个时间点的异常; 并根据每个时间点的计算出的异常来检测设备的异常周期。使用设备状态监视方法的设备被设计为在预先存储的学习数据之间选择接近观察数据的预定数量的学习数据，通过使用所选择的学习数据创建正常模型，并且计算观测数据的异常通过使用观察数据和正常模型。

2. 发明专利

JP2013025367A Facility state monitoring method and device of the same 审中-公开
标题翻译：设备状态监控方法及其设备
公开(公告)号：JP2013025367A
公开(公告)日：2013-02-04
申请号：JP2011156648
申请日：2011-07-15
申请人： Wakayama Univ , 国立大学法人和歌山大学 , Hitachi Ltd , 株式会社日立製作所
发明人： SHIBUYA HISAE , MAEDA SHUNJI , WADA TOSHIKAZU , OZAKI SHINSAKU
IPC分类号： G05B23/02 , G06Q50/10
CPC分类号： G05B23/024 , G06Q50/10
摘要： PROBLEM TO BE SOLVED: To solve the problems in which a lot of learning data is required in order to improve accuracy of a regression model and calculation time becomes long, while, in abnormality detection in a facility of a plant or the like, the regression model is prepared from normal input and output data, a prediction is made on the basis of observation data and the regression model and whether or not it is abnormal is discriminated on the basis of a prediction error.SOLUTION: The regression model is prepared using a prescribed number of pieces of data close to observation data from stored learning data, an abnormality degree of the observation data is calculated on the basis of the regression model, and whether or not it is abnormal is discriminated by comparing it with a threshold. For the data not determined as being abnormal, similarity with the closest learning data is calculated, the learning data is replaced when the similarity is high, and it is added to the learning data when the similarity is low.
摘要翻译：要解决的问题为了解决为了提高回归模型的精度和计算时间而需要大量学习数据的问题，在植物等的设备的异常检测中，，从正常输入和输出数据准备回归模型，基于观测数据和回归模型进行预测，并根据预测误差来判别是否异常。
解决方案：使用接近来自存储的学习数据的观察数据的规定数量的数据准备回归模型，基于回归模型计算观测数据的异常程度，以及是否是通过将其与阈值进行比较来区分异常。对于未被确定为异常的数据，计算与最接近的学习数据的相似度，当相似度高时，学习数据被替换，并且当相似度较低时被添加到学习数据。版权所有（C）2013，JPO＆INPIT

3. 发明专利

JP2013152655A Abnormality diagnostic method and health management method for plant or facility 审中-公开
标题翻译：异常诊断方法和工厂或设施的健康管理方法
公开(公告)号：JP2013152655A
公开(公告)日：2013-08-08
申请号：JP2012013689
申请日：2012-01-26
申请人： Hitachi Ltd , 株式会社日立製作所
发明人： MAEDA SHUNJI , SHIBUYA HISAE
IPC分类号： G05B23/02
CPC分类号： G05B23/0283
摘要： PROBLEM TO BE SOLVED: To provide a method for presuming a progress of abnormality degree from sensor data, operation information, event information, facility load, a work report, and the like, grasping the health status of a plant or facility, and estimating an operation continuation validation time, for maintaining and improving the availability of the plant or facility.SOLUTION: A method for predicting an abnormality degree with high accuracy and predicting an operation continuation validation time consists of the steps of: (1) predicting an abnormality measure 1 by using the Gaussian process which is a nonlinear regression method; (2) predicting an abnormality measure by applying a recognition technique such as k-NN to time series data; (3) modeling dynamics for time series data 7 to obtain a state space model and predicting the abnormality measure 1 by a particle filter; (4) predicting RUL from the abnormality measure; (5) modeling dynamics to obtain a state space model for the abnormality measure and also for operation information, event information, and facility load, and predicting the operation continuation validation time by the particle filter; and the like.
摘要翻译：要解决的问题：提供一种用于从传感器数据，操作信息，事件信息，设施负载，工作报告等推测异常程度的进展的方法，掌握设备或设施的健康状况，以及估计操作延续验证时间，用于维护和改善工厂或设施的可用性。解决方案：一种以高精度预测异常程度并预测操作连续验证时间的方法包括以下步骤：（1）预测异常测量1 通过使用作为非线性回归方法的高斯过程; （2）通过对时间序列数据应用诸如k-NN的识别技术来预测异常测量; （3）时间序列数据7的建模动态，以获得状态空间模型并通过粒子滤波器预测异常测量1; （4）从异常测量中预测RUL; （5）模拟动力学，以获得异常测量的状态空间模型，以及操作信息，事件信息和设施负载，以及通过粒子滤波器预测操作持续验证时间; 等等。

4. 发明专利

JP2011227706A Abnormality detection and diagnosis method, abnormality detection and diagnosis system, and abnormality detection and diagnosis program 有权
标题翻译：异常检测和诊断方法，异常检测和诊断系统，异常检测和诊断程序
公开(公告)号：JP2011227706A
公开(公告)日：2011-11-10
申请号：JP2010096873
申请日：2010-04-20
申请人： Hitachi Ltd , 株式会社日立製作所
发明人： MAEDA SHUNJI , SHIBUYA HISAE , MAGARA HIROYUKI
IPC分类号： G05B23/02 , F01D25/00 , F02C7/00 , F02C9/00 , G01M99/00 , G06Q50/00 , G06Q50/10
CPC分类号： G06F15/00 , G05B23/0224 , G06Q10/06
摘要： PROBLEM TO BE SOLVED: To provide an abnormality detection and diagnosis method and system that can detect an abnormality at an early stage and with high sensitivity in a facility such as a plant.SOLUTION: Maintenance history information composed of past events such as work history and part replacement information is associated with each other on a keyword basis. An abnormality is detected based on abnormality detection targeting an output signal of a multidimensional sensor installed in a facility, and the detected abnormality is connected to associated maintenance history information, whereby a diagnosis and measures necessary for the occurred abnormality are determined.
摘要翻译：要解决的问题：提供能够在诸如植物的设施中早期检测异常并且具有高灵敏度的异常检测和诊断方法和系统。
解决方案：由过去事件（如工作历史和部分替换信息）组成的维护历史信息在关键字的基础上相互关联。基于针对安装在设施中的多维传感器的输出信号的异常检测来检测异常，并且将检测到的异常与相关的维护历史信息相关联，由此确定发生的异常所需的诊断和措施。版权所有（C）2012，JPO＆INPIT

5. 发明专利

JP2003329608A Inspection condition judgement program, inspection apparatus and inspection system 审中-公开
标题翻译：检验条件审查程序，检验装置和检查系统
公开(公告)号：JP2003329608A
公开(公告)日：2003-11-19
申请号：JP2002132260
申请日：2002-05-08
申请人： Hitachi Ltd , 株式会社日立製作所
发明人： ONO MAKOTO , ASAKAWA YOHEI , IWATA HISAFUMI , HARADA KANAKO , TAKAGI YUJI , SHIBUYA HISAE
IPC分类号： G01N21/956 , G06T1/00 , H01L21/66
摘要： PROBLEM TO BE SOLVED: To judge whether the inspection condition of an inspection apparatus is appropriate or not by using a computer equipped in the inspection apparatus and an inspection system of products which forms and manufactures a plurality of products such as a semiconductor integrated circuit on a substrate at the same time. SOLUTION: A defect coordinate data are input by a defect coordinate data input process 11. A conditional branch 12 with the number of defects and a first threshold value is performed. When the number of defects is larger than the first threshold value, a conditional branch 15 is performed with a dispersion and a second threshold value after performing a defect number calculation process 13 according to areas in a chip and a dispersion calculation process 14. When the dispersion is larger than the second threshold value, a warning process 16 is performed. On the other hand, a warning process 17 is performed when the number of defects is the first threshold value or below. COPYRIGHT: (C)2004,JPO
摘要翻译：要解决的问题：通过使用检查装置中配备的计算机以及形成和制造诸如半导体集成的多个产品的产品的检查系统来判断检查装置的检查条件是否适合电路同时在基板上。解决方案：缺陷坐标数据由缺陷坐标数据输入处理11输入。执行具有缺陷数量和第一阈值的条件分支12。当缺陷数大于第一阈值时，根据芯片和色散计算处理14中的区域执行缺陷数计算处理13之后，利用色散和第二阈值执行条件分支15.当色散大于第二阈值，则执行警告处理16。另一方面，当缺陷的数量为第一阈值或更低时执行警告处理17。版权所有（C）2004，JPO

6. 发明专利

JP2012181209A Defect classification method and apparatus therefor 有权
标题翻译：缺陷分类方法及其设备
公开(公告)号：JP2012181209A
公开(公告)日：2012-09-20
申请号：JP2012134507
申请日：2012-06-14
申请人： Hitachi Ltd , 株式会社日立製作所
发明人： SHIBUYA HISAE , MAEDA SHUNJI , HAMAMATSU REI
IPC分类号： G01N21/956 , H01L21/66
摘要： PROBLEM TO BE SOLVED: To solve such a problem that in defect classification of appearance examination, in spite of the needs of adjusting the purity or the accuracy or the both of an important defect to be target values or more, since teaching-type defect classification is conditionally set to improve a classification accuracy rate, it is impossible to respond to such needs.SOLUTION: A defect classification apparatus includes a feature amount extraction unit, a defect classification unit and a classification condition setting unit, the classification condition setting unit has a function for correspondingly teaching a feature amount of a defect and a class of accuracy and a function for designating the priority order of classification, and conditional setting is performed so as to improve an accuracy rate for the classification of high priority.
摘要翻译：要解决的问题：为了解决在外观检查的缺陷分类中的问题，尽管需要将纯度或精度或重要缺陷的两个调整为目标值以上，有条件地设置类型缺陷分类以提高分类准确率，不可能满足这种需求。解决方案：缺陷分类装置包括特征量提取单元，缺陷分类单元和分类条件设置单元，分类条件设置单元具有相应地教导缺陷的特征量和精度等级的功能，执行用于指定分类的优先级顺序的功能，并且执行条件设置，以便提高高优先级分类的准确率。版权所有（C）2012，JPO＆INPIT

7. 发明专利

JP2010092355A Abnormality detection method and system 有权
标题翻译：异常检测方法与系统
公开(公告)号：JP2010092355A
公开(公告)日：2010-04-22
申请号：JP2008263030
申请日：2008-10-09
申请人： Hitachi Ltd , 株式会社日立製作所
发明人： MAEDA SHUNJI , SHIBUYA HISAE
IPC分类号： G05B23/02 , G06Q50/00 , G06Q50/10
CPC分类号： G05B23/0254 , G06K9/00536 , G06K9/6252 , G06K9/6272 , G06K9/6284
摘要： PROBLEM TO BE SOLVED: To provide a method and system for tolerating incompleteness of learning data and mixing of abnormalities and enabling highly precise discovery in an early stage of the abnormality in equipment such as a plant. SOLUTION: In order to attain the objective, (1) a locus is divided into clusters in chase of time by paying attention on behavior of temporal data. (2) Modeling of a divided cluster group is carried out in a subspace and an outlier is computed as an abnormality candidate. (3) A state transition by change with time, environmental variation, maintenance (parts replacement), and a working condition is obtained by using the learning data as a reference (compare, reference, or the like). (4) Modeling shall be based on a subspace method, such as a regression analysis method with omission of N piece of data (N=0, 1, 2, and so on) or a projection distance method (for example, when N=1, modeling is carried out by excluding this, assuming that one abnormal data is mixed) or a local subspace method. Meanwhile applying of a straight line in a regression analysis method corresponds to the regression analysis of the lowest degree. COPYRIGHT: (C)2010,JPO&INPIT
摘要翻译：要解决的问题：提供一种方法和系统，用于容忍学习数据的不完备性和异常混合，并能够在诸如设备的设备的异常的早期阶段中进行高度精确的发现。解决方案：为了达到目的，（1）通过注意时间数据的行为，将时间段划分成一个时间段。（2）分割的群组的建模在子空间中进行，异常值被计算为异常候选。（3）通过使用学习数据作为参考（比较，参考等），获得随时间，环境变化，维护（部件更换）和工作状态的状态转变。（4）建模应基于子空间方法，例如省略N条数据（N = 0,1,2等）或投影距离法的回归分析方法（例如，当N = 1，假设一个异常数据混合进行排除，或者使用本地子空间方法进行建模。同时在回归分析方法中应用直线对应于最低程度的回归分析。版权所有（C）2010，JPO＆INPIT

8. 发明专利

JP2004191187A Method and apparatus for analyzing defect composition 有权
公开(公告)号：JP2004191187A
公开(公告)日：2004-07-08
申请号：JP2002359577
申请日：2002-12-11
申请人： Hitachi Ltd , 株式会社日立製作所
发明人： OBARA KENJI , TAKAGI YUJI , SHIBUYA HISAE , HOSOYA NAOKI
IPC分类号： G01N23/225 , G01Q30/02 , G01R31/26 , G06T7/00 , H01J37/147 , H01J37/252 , H01L21/66
CPC分类号： G06T7/0004 , G06T2207/30148
摘要： PROBLEM TO BE SOLVED: To analyze a defect composition in a highly precise and efficient manner by detecting the irradiation position of an electron beam coinciding with a defect position. SOLUTION: When the defect composition to be analyzed is sorted (step 114) and the irradiation condition of the electron beam is set for the purpose of EDX analysis (step 115), the electron beam is used at a defect corresponding position corresponding to the position of the defect of a chip in the vicinity of a target chip including a defect to acquire a low magnification reference image (6) of low magnification (steps 116 and 117). Next, a low magnification defect image (7) of the same low magnification at the defect position of the target chip is acquired (steps 118 and 119). The defect position is acquired (step 120) by comparing these low magnification images (6) and (7) with each other and the electron beam is deflected to irradiate the defect position to acquire the composition spectrum of the defect (step 121). COPYRIGHT: (C)2004,JPO&NCIPI

9. 发明专利

JP2004153228A Method for analyzing defect information and its apparatus 有权
公开(公告)号：JP2004153228A
公开(公告)日：2004-05-27
申请号：JP2003037596
申请日：2003-02-17
申请人： Hitachi Ltd , 株式会社日立製作所
发明人： OBARA KENJI , TAKAGI YUJI , SHIBUYA HISAE
IPC分类号： H01L21/66
摘要： PROBLEM TO BE SOLVED: To provide an apparatus for effectively extracting equipment in which a trouble has occurred, in a manufacturing process of thin film devices. SOLUTION: In the manufacturing process of thin film devices, the apparatus extracts the candidates of manufacturing equipment in which a trouble has occurred, by evaluating the relationship between data obtained in relation to product inspection for a product, and data indicating the state of manufacturing equipment, for effectively extracting equipment in which a trouble has occurred. COPYRIGHT: (C)2004,JPO

10. 发明专利

JP2014149840A Abnormality detection method and system therefor 有权
标题翻译：异常检测方法及其系统
公开(公告)号：JP2014149840A
公开(公告)日：2014-08-21
申请号：JP2014048365
申请日：2014-03-12
申请人： Hitachi Ltd , 株式会社日立製作所
发明人： MAEDA SHUNJI , SHIBUYA HISAE
IPC分类号： G05B23/02 , G06Q50/04 , G06Q50/06
CPC分类号： Y02P90/30
摘要： PROBLEM TO BE SOLVED: To detect, even when a plurality of abnormalities occur at the same time or at short time intervals in a facility such as a plant, and even when those abnormalities are different types, those abnormalities or their signs with high sensitivity and in an early stage.SOLUTION: An abnormality detection method includes: acquiring data related to an operational status from a plurality of sensors installed in a plant or a facility; modeling learning data corresponding to almost normal data in a normal operational status; calculating the anomaly measure of the data acquired from the plurality of sensors by using the modeled learning data; and performing abnormality detection on the basis of the calculated anomaly measure. In a process of calculating the anomaly measure, a residual is calculated from the modeled learning data about the data acquired from the plurality of sensors, and a signal having the residual larger than a predetermined value is removed, and the anomaly measure is recursively calculated about the data acquired from the plurality of sensors from which the signal having the large residual has been removed so that it is possible to achieve abnormality detection.
摘要翻译：要解决的问题：即使当在诸如植物的设施中同时或以短时间间隔发生多个异常时，即使当这些异常是不同类型的那些异常或其具有高灵敏度的迹象，以及解决方案：异常检测方法包括：从安装在工厂或设施中的多个传感器获取与操作状态有关的数据; 在正常操作状态下建模与几乎正常数据对应的学习数据; 通过使用建模的学习数据计算从多个传感器获取的数据的异常测量; 并且基于所计算的异常测量来执行异常检测。在计算异常测量的过程中，从关于从多个传感器获取的数据的建模学习数据计算残差，并且消除具有大于预定值的残差的信号，并且递归地计算异常测量从具有大残留的信号的多个传感器获取的数据被去除，使得可以实现异常检测。

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式