专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

11. 发明专利

JP2013140495A Plant state monitoring device and plant state monitoring method 有权
标题翻译：植物状态监测装置和植物状态监测方法
公开(公告)号：JP2013140495A
公开(公告)日：2013-07-18
申请号：JP2012000350
申请日：2012-01-05
申请人： Hitachi Ltd , 株式会社日立製作所
发明人： SEKIAI TAKAO , EGUCHI TORU , KUSUMI HISAHIRO , FUKAI MASAYUKI , SHIMIZU SATORU , MURAKAMI MASAHIRO
IPC分类号： G05B23/02
摘要： PROBLEM TO BE SOLVED: To provide a plant state monitoring device that negates the need of prior knowledge with respect to a plant that is an application object and can automatically acquire information contributing to a reduction in operating costs.SOLUTION: A plant state monitoring device for monitoring an operational state of a plant 100 comprises: model data selection means 400 for determining a given period of time before and after time when a maximum value is obtained or a given period of time before and after time when a minimum value is obtained as a period of time of model data for at least one time-series data of time-series data of efficiency of the plant 100 calculated from a measurement signal 4 and time-series data of the measurement signal 4 for an environmental load substance and for selecting the time-series data in the period of time of the model data as the model data; and state monitor means 500 that classifies the model data into a plurality of categories, constructs monitor models 600A through 600C, determines whether or not a current measurement signal 6 belongs to the plurality of categories of the monitor models 600A through 600C and outputs a determination result to an image display device 940.
摘要翻译：要解决的问题：提供一种植物状态监测装置，其不需要相对于作为应用对象的植物的先验知识，并且可以自动获取有助于降低操作成本的信息。解决方案：用于监视工厂100的操作状态包括：模型数据选择装置400，用于确定获得最大值时间之前和之后的给定时间段，或者当获得最小值时间之前和之后的给定时间段作为从测量信号4计算的工厂100的效率的时间序列数据的至少一个时间序列数据的模型数据的时间段和环境负荷物质的测量信号4的时间序列数据，以及用于选择时间序列数据在模型数据的时间段作为模型数据; 以及将模型数据分类为多个类别的状态监视装置500，构成监视模型600A〜600C，判断当前测量信号6是否属于监视模型600A〜600C的多个类别，并输出确定结果到图像显示装置940。

12. 发明专利

JP2013061695A Power-generating plant diagnosis device, and power-generating plant diagnosis method 有权
标题翻译：动力发电厂诊断装置和发电厂诊断方法
公开(公告)号：JP2013061695A
公开(公告)日：2013-04-04
申请号：JP2011197837
申请日：2011-09-12
申请人： Hitachi Ltd , 株式会社日立製作所
发明人： SEKIAI TAKAO , EGUCHI TORU , KUSUMI HISAHIRO , FUKAI MASAYUKI , SHIMIZU SATORU , MURAKAMI MASAHIRO
IPC分类号： G05B23/02
摘要： PROBLEM TO BE SOLVED: To provide a diagnosis device which reduces a false report that cannot detect abnormality when the abnormality occurs, and has high diagnostic accuracy.SOLUTION: A power-generating plant diagnosis device diagnoses an operation state of a plant on the basis of a measurement signal for measuring state amount from a power-generating plant, and indicates a diagnosis result on an image display device. The power-generating plant diagnosis device includes: model constructing means 700 for constructing a model to be used for diagnosis by using the measurement signal for measuring the state amount of the power-generating plant; model defining means 400 for defining a normalization method of the operation condition diagnosed by the above-mentioned model and the measurement signal; and diagnosing means 800 for diagnosing the operation state of the power-generating plant by using the model constructed by the model constructing means. The model defining means includes an operation condition determining portion 500 for determining the operation conditions of the power-generating plant; and a normalization condition deciding portion 600 for deciding the normalization condition of data about each operation condition determined by the operation condition determining portion. The diagnosing means diagnoses the operation state by switching the diagnosis model according to the operation condition.
摘要翻译：要解决的问题：提供一种诊断装置，其减少当异常发生时不能检测到异常的虚假报告，并且具有高诊断精度。解决方案：发电厂诊断装置基于用于测量来自发电厂的状态量的测量信号来诊断工厂的操作状态，并且在图像显示装置上指示诊断结果。发电厂诊断装置包括：模型构成单元700，用于通过使用测量信号构造用于诊断的模型，用于测量发电厂的状态量; 用于定义由上述模型诊断的操作条件和测量信号的归一化方法的模型定义装置400; 以及诊断装置800，用于通过使用由模型构建装置构造的模型来诊断发电厂的运行状态。模型定义装置包括用于确定发电厂的操作条件的操作条件确定部分500; 以及归一化条件决定部600，用于判定由操作条件确定部确定的关于每个操作条件的数据的归一化条件。诊断装置根据操作条件切换诊断模型来诊断操作状态。版权所有（C）2013，JPO＆INPIT

13. 发明专利

JP2013047970A Maintenance tool and maintenance tool of controller 有权
标题翻译：控制器的维护工具和维护工具
公开(公告)号：JP2013047970A
公开(公告)日：2013-03-07
申请号：JP2012229846
申请日：2012-10-17
申请人： Hitachi Ltd , 株式会社日立製作所
发明人： EGUCHI TORU , YAMADA AKIHIKO , SEKIAI TAKAO , FUKAI MASAYUKI , SHIMIZU SATORU
IPC分类号： G05B13/04 , G05B13/02
摘要： PROBLEM TO BE SOLVED: To provide a maintenance tool capable of constructing a highly accurate statistical model and acquiring a desired control effect even when controlling a learning type plant with bias in data used when constructing the statistical model.SOLUTION: The maintenance tool of the present invention is configured to display transition of a compressional degree that changes relative to the number m of times of repetition of model adjustment, on the basis of information on the number of times of the repetition of the model adjustment preserved in a model construction database 220 and information on the change of the compressional degree.
摘要翻译：要解决的问题：提供一种能够构建高精度的统计模型并获得期望的控制效果的维护工具，即使在构建统计模型时使用的数据中控制具有偏差的学习型工厂。解决方案：本发明的维护工具被配置为基于关于重复次数的信息来显示相对于模型调整的重复次数m变化的压缩程度的转变在模型构造数据库220中保留的模型调整和关于压缩程度的变化的信息。版权所有（C）2013，JPO＆INPIT

14. 发明专利

JP2012141862A Control device for plant and control device for fire power generation plant 有权
标题翻译：用于火力发电厂的设备和控制装置的控制装置
公开(公告)号：JP2012141862A
公开(公告)日：2012-07-26
申请号：JP2011000301
申请日：2011-01-05
申请人： Hitachi Ltd , 株式会社日立製作所
发明人： EGUCHI TORU , KUSUMI HISAHIRO , SEKIAI TAKAO , FUKAI MASAYUKI , SHIMIZU SATORU
IPC分类号： G05B13/02 , G05B13/04
CPC分类号： G05B13/048
摘要： PROBLEM TO BE SOLVED: To provide a control device of a plant for, when adjusting a statistical model by on-line via an RBF network by using the measurement data of a plant, improving the estimating accuracy of the statistical model by ending optimized calculation within a control period.SOLUTION: The control device of a plant includes: a statistical model for estimating the value of a measurement signal to be acquired when applying a control signal to the plant; a model construction database for storing data to be used for the construction of the statistical model; a data pre-processing part for generating model construction data by operating statistical processing to the measurement signal; an operation method learning part for learning the generation method of a model input so that a target value can be achieved by a model output; and a model adjustment part for adjusting the radial parameter of the statistical model included in information stored in the model construction database. The statistical model is configured to generate the model output by using the result of adjustment of the radial parameter by the model adjustment part .
摘要翻译：要解决的问题：为了提供一种工厂的控制装置，当通过使用工厂的测量数据通过RBF网络在线调整统计模型时，通过结束来提高统计模型的估计精度在控制期内优化计算。解决方案：工厂的控制装置包括：用于估计当向工厂应用控制信号时要获取的测量信号的值的统计模型; 用于存储要用于构建统计模型的数据的模型构建数据库; 数据预处理部，其通过对所述测量信号进行统计处理来生成模型构造数据; 用于学习模型输入的生成方法的操作方法学习部分，从而可以通过模型输出来实现目标值; 以及用于调整包含在模型构造数据库中存储的信息中的统计模型的径向参数的模型调整部分。统计模型配置为通过模型调整部分使用径向参数的调整结果生成模型输出。版权所有（C）2012，JPO＆INPIT

15. 发明专利

JP2011070334A Abnormality diagnostic device and abnormality diagnostic method 有权
标题翻译：异常诊断装置和非正常诊断方法
公开(公告)号：JP2011070334A
公开(公告)日：2011-04-07
申请号：JP2009219901
申请日：2009-09-25
申请人： Hitachi Ltd , 株式会社日立製作所
发明人： SEKIAI TAKAO , EGUCHI TORU , OKAZAKI TERUYUKI , SHIMIZU SATORU , FUKAI MASAYUKI
IPC分类号： G05B23/02
摘要： PROBLEM TO BE SOLVED: To provide an abnormality diagnostic device capable of appropriately adjusting a resolution parameter, and highly precisely detecting abnormality, the abnormality diagnostic device using a clustering technology. SOLUTION: A reference signal determination means has a function for dividing a measurement signal in a normal time of a plant into two or more data groups, and a classification means includes: a learning part for classifying one of the data groups into categories: a test part for classifying one of the residual data groups, which are not used by the learning part, into categories; and a resolution parameter adjustment part for adjusting a resolution parameter to be used when classifying the data group into categories, by using the classification results of the learning part and the test part. COPYRIGHT: (C)2011,JPO&INPIT
摘要翻译：解决问题的方法为了提供能够适当地调整分辨率参数，高精度地检测异常的异常诊断装置，使用聚类技术进行异常诊断。解决方案：参考信号确定装置具有将工厂的正常时间中的测量信号分成两个或更多个数据组的功能，并且分类装置包括：学习部分，用于将数据组中的一个分类成类别：用于将学习部分未使用的剩余数据组之一分类为分类的测试部件; 以及分辨率参数调整部，其通过使用所述学习部和所述测试部的分类结果来调整将所述数据组分类成类别时所使用的分辨率参数。版权所有（C）2011，JPO＆INPIT

16. 发明专利

JP2010237893A Device and method for diagnosing abnormality of plant 有权
标题翻译：用于诊断植物异常的装置和方法
公开(公告)号：JP2010237893A
公开(公告)日：2010-10-21
申请号：JP2009083989
申请日：2009-03-31
申请人： Hitachi Ltd , 株式会社日立製作所
发明人： SEKIAI TAKAO , YAMADA AKIHIKO , EGUCHI TORU , KUSUMI HISAHIRO , HORI YOSHINARI , FUKAI MASAYUKI , SHIMIZU SATORU
IPC分类号： G05B23/02
摘要： PROBLEM TO BE SOLVED: To provide an abnormality diagnosing device for a plant capable of reducing the risk of erroneous diagnosis. SOLUTION: This diagnosing device 200 includes a measured signal database 230 storing a measured signal of the plant, an abnormality determination database 250 storing category classification information with normal-time or abnormal-time data classified into categories, and the appearance frequency of the category classification information, and a comparison means 700 for comparing a result, of the data stored in the measured signal database 230, classified into the category, with the category classification information stored in the abnormality determination database 250, using information of the appearance frequency of the category classification information, so as to diagnose an operation state of the plant, and for outputting information of announcing abnormality, as an image display data, when a diagnostic result is abnormal. COPYRIGHT: (C)2011,JPO&INPIT
摘要翻译：要解决的问题：提供一种能够降低错误诊断风险的设备的异常诊断装置。解决方案：该诊断装置200包括存储工厂的测量信号的测量信号数据库230，存储分类为正常时间或异常时间数据的类别分类信息的异常判定数据库250，类别分类信息，以及比较装置700，用于将存储在测量信号数据库230中的分类到类别的数据与存储在异常判定数据库250中的类别分类信息的结果，使用出现频率的信息进行比较的诊断结果异常的情况下，诊断植物的运转状态，输出通知异常的信息作为图像显示数据。版权所有（C）2011，JPO＆INPIT

17. 发明专利

JP2010146068A Control device for plant, and control device of thermal power generation plant 审中-公开
标题翻译：用于植物的控制装置和热发电厂的控制装置
公开(公告)号：JP2010146068A
公开(公告)日：2010-07-01
申请号：JP2008319568
申请日：2008-12-16
申请人： Hitachi Ltd , 株式会社日立製作所
发明人： EGUCHI TORU , YAMADA AKIHIKO , KUSUMI HISAHIRO , SEKIAI TAKAO , FUKAI MASAYUKI , SHIMIZU SATORU , HORI YOSHINARI
IPC分类号： G05B13/02 , G05B13/04 , G06N3/00
摘要： PROBLEM TO BE SOLVED: To provide a control device of a plant for learning a desired operation method within a control cycle by using a learning policy for efficiently solving a problem autonomously extracted by a retrieval means included in a control device, in controlling a plant in which the characteristics of a control object are complicated, and information is imperfect. SOLUTION: The control device is provided with: a model for estimating a value of a measurement signal to be acquired when a control signal is applied to a plant; an operation method learning part for learning a method of generating a model input so that the model output can achieve a target value; a learning information database for storing information related with the constraints and learning results of learning in an operation method learning part; an optimal solution retrieval part for retrieving the model input equivalent to operation conditions from which a desired model output is obtained; and a retrieval information database for storing information related with the constraints of retrieval and the optimal solution retrieval in the optimal solution retrieval part. An operation method learning part generates a model input by using information related with the optimal operation conditions obtained by the optimal solution retrieval part. COPYRIGHT: (C)2010,JPO&INPIT
摘要翻译：要解决的问题：为了提供一种用于在控制周期内学习所需操作方法的设备的控制装置，通过使用学习策略来有效地解决由控制装置中包括的检索装置自主提取的问题，控制控制对象的特征复杂，信息不完善的植物。解决方案：控制装置设置有：当将控制信号施加到工厂时估计要获取的测量信号的值的模型; 操作方法学习部，用于学习生成模型输入的方法，使得模型输出可以实现目标值; 用于存储与操作方法学习部分中的学习的约束和学习结果有关的信息的学习信息数据库; 用于检索等效于获得期望模型输出的操作条件的模型输入的最优解求解部分; 以及检索信息数据库，用于存储与最优解检索部分中的检索约束和最优解方案检索相关的信息。操作方法学习部分通过使用与由最优解检索部分获得的最佳操作条件相关的信息来生成模型输入。版权所有（C）2010，JPO＆INPIT

18. 发明专利

JP2009198136A Gas concentration estimation device and gas concentration estimation method for coal-fired boiler 有权
标题翻译：煤气浓度估算装置和燃煤浓度估算方法
公开(公告)号：JP2009198136A
公开(公告)日：2009-09-03
申请号：JP2008042796
申请日：2008-02-25
申请人： Hitachi Ltd , 株式会社日立製作所
发明人： HAYASHI YOSHIHARU , YAMADA AKIHIKO , SEKIAI TAKAO , FUKAI MASAYUKI , SHIMIZU SATORU
IPC分类号： F22B35/18 , F23C99/00 , F23N5/00 , G08B21/14
摘要： PROBLEM TO BE SOLVED: To provide a gas concentration estimation device for a coal-fired boiler accurately estimating the concentration of a gas component by suppressing an estimation error caused by a measurement error included in actual unit data when simulating change of the gas component in exhaust gas of the coal-fired boiler by using a neural network.
SOLUTION: The gas concentration estimation device estimating the concentration of the gas component in the exhaust gas discharge from the coal-fired boiler by using the neural network is composed of a process database part storing process data of the boiler, a filtering process part carrying out a filtering process of extracting data suiting learning of the neural network from the stored process data, a neural network learning process part carrying out a learning process of the neural network on the basis of the extracted data suiting learning of the neural network, and a neural network estimation process part estimating a CO concentration of an NOx concentration in the exhaust gas discharged from the boiler on the basis of the learning process.
COPYRIGHT: (C)2009,JPO&INPIT
摘要翻译：要解决的问题：为了提供燃煤锅炉的气体浓度估计装置，通过在模拟气体变化时抑制由实际单位数据中包括的测量误差引起的估计误差，精确地估计气体成分的浓度通过使用神经网络对燃煤锅炉废气中的组分进行分析。解决方案：使用神经网络估计来自燃煤锅炉的废气排放中的气体成分浓度的气体浓度估计装置由存储锅炉的过程数据的过程数据库部分，过滤处理执行从存储的过程数据提取适合于神经网络的学习的数据的过滤过程的部分，神经网络学习处理部分，基于提取的神经网络的学习数据，执行神经网络的学习过程，以及神经网络估计处理部，其基于所述学习处理来估计从所述锅炉排出的废气中的NOx浓度的CO浓度。版权所有（C）2009，JPO＆INPIT

19. 发明专利

JP2009110256A Plant control device, and plant control device for thermal power plant 有权
标题翻译：植物控制装置和用于热力发电厂的植物控制装置
公开(公告)号：JP2009110256A
公开(公告)日：2009-05-21
申请号：JP2007281762
申请日：2007-10-30
申请人： Hitachi Ltd , 株式会社日立製作所
发明人： EGUCHI TORU , YAMADA AKIHIKO , SEKIAI TAKAO , FUKAI MASAYUKI , SHIMIZU SATORU
IPC分类号： G05B13/02 , G05B13/04
摘要： PROBLEM TO BE SOLVED: To provide a plant control device capable of learning within a control period regardless of the number of model inputs and outputs.
SOLUTION: The plant control device divides model input and output for simulating plant behavior, speeds up the learning by learning an operation method of the model input divided into a plurality of learning means, and thereby enables the learning to be finished within the control period. The plant control device includes: a measurement signal database; an operation signal database; a model for estimating a measurement signal value when the plant is provided with the operation signal; the plurality of learning means which divide the model input and the model output, respectively corresponding to the operation signal and the measurement signal, into a plurality of groups, and learn a method for generating the model input of each group so that the model output of each group achieve a target set value; a function for aggregating model inputs of each group generated by the learning means and inputting them into the model; and a model input/output generation means for dividing the model output according to division setting information of the model output of each group, and functioning to output the divided model outputs to the corresponding learning means.
COPYRIGHT: (C)2009,JPO&INPIT
摘要翻译：要解决的问题：提供能够在控制周期内学习的工厂控制装置，而与模型输入和输出的数量无关。解决方案：工厂控制装置将模型输入和输出分为模拟植物行为，通过学习分为多个学习手段的模型输入的操作方法加快学习速度，从而使学习能够在控制期工厂控制装置包括：测量信号数据库; 操作信号数据库; 用于在所述设备被提供有所述操作信号时估计测量信号值的模型; 将分别对应于操作信号和测量信号的模型输入和模型输出分成多个组的多个学习装置，并且学习用于生成每个组的模型输入的方法，使得模型输出每组达到目标设定值; 用于聚合由学习装置产生的每组的模型输入并将其输入到模型中的功能; 以及模型输入/输出生成装置，用于根据每组的模型输出的分割设置信息对模型输出进行分割，并且将分割的模型输出输出到相应的学习装置。版权所有（C）2009，JPO＆INPIT

20. 发明专利

JP2008076023A Plant control device 有权
标题翻译：植物控制装置
公开(公告)号：JP2008076023A
公开(公告)日：2008-04-03
申请号：JP2006259053
申请日：2006-09-25
申请人： Hitachi Ltd , 株式会社日立製作所
发明人： KANEDA MASAKI , YAMADA AKIHIKO , SEKIAI TAKAO , HAYASHI YOSHIHARU , FUKAI MASAYUKI , SHIMIZU SATORU
IPC分类号： F22B35/18
摘要： PROBLEM TO BE SOLVED: To provide a plant control technology superior in performance, by reflecting an operation time boiler shape on construction of a control model.
SOLUTION: This plant control device has a boiler shape analyzing means 2 acquiring data by measuring the boiler shape when operating a boiler, a measuring information processing means 3 calculating a parameter for numerically analyzing a physical phenomenon of the boiler on the basis of the data, an analytical condition setting means 4 setting an analytical condition by using the calculated parameter, a numerical analytical means 5 numerically analyzing a reaction phenomenon of a plant to a plurality of operation conditions in response to the preset analytical condition, a control model setting means 7 setting a control model by performing strengthening learning by using a made continuous model by making the continuous model approximate so that a process value in the respective operation conditions continues to a change in respective operation quantities on the basis of an acquired analytical result, and an operation quantity arithmetic operation means 9 arithmetically operating the operation quantity to the plant by using the preset control model and operation data acquired from the plant.
COPYRIGHT: (C)2008,JPO&INPIT
摘要翻译：要解决的问题：为了提供性能优越的工厂控制技术，通过反映操作时间锅炉形状对控制模型的构建。解决方案：该设备控制装置具有锅炉形状分析装置2，其在操作锅炉时通过测量锅炉形状来获取数据;测量信息处理装置3，基于以下步骤计算用于数值分析锅炉物理现象的参数数据，分析条件设定装置4，通过使用计算出的参数来设定分析条件;数值分析装置5，其根据预设的分析条件对植物的反应现象进行多个操作条件的数值分析;控制模型设定通过使连续模型近似，通过使用连续模型进行加强学习，使得各个操作条件中的过程值基于所获得的分析结果继续改变各个操作量，来设置控制模型7;以及算术运算装置9 通过使用从工厂获取的预设控制模型和操作数据对工厂的操作量。版权所有（C）2008，JPO＆INPIT

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