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

JPH1144440A WATER HEAT STORAGE TANK AND WATER HEAT STORAGE DEVICE 失效
公开(公告)号：JPH1144440A
公开(公告)日：1999-02-16
申请号：JP21586597
申请日：1997-07-24
申请人： KAWASAKI HEAVY IND LTD , TOKYO ELECTRIC POWER CO
发明人： MASUYAMA KAZUO , TANAKA HIDEO , TAJIRI HIRONORI , SHIBATA SOICHIRO , HAMAYASHIKI TAKESHI
IPC分类号： F24F5/00
摘要： PROBLEM TO BE SOLVED: To permit the contriving of cost down as well as the formation of a sufficient temperature stratification and realize a high heat storage efficiency by a method wherein the tank main body of a sealed and pressure resistant structure is formed by connecting lids, made of a steal sheet, to the upper end as well as the lower end of a continuous spiral steel pipe while a distributor and a high-temperature side water flow passing pipe are incorporated into the tank main body. SOLUTION: A water heat storage tank 21, employed for the air conditioning and the like of a large-scale factory and the like, capable of being installed at an outdoor vacant land, is provided with a tank main body 20 of a sealed and pressure resistant structure, in which lids 16, 18, made by a steel sheet, are attached to the upper and lower ends of a continuous and large diametral spiral steel pipe. In this case, an upper distributor 22 and a lower distributor 24 are provided at upper and lower parts in the tank main body to constitute a temperature stratification at the upper and lower parts in the tank main body 20 and keep a high heat storage efficiency while the distributors 22, 24 are formed by providing a plurality of sets of openings 26 on both left and right sides of a horizontal tube in the lengthwise direction of the same. On the other hand, low-temperature side water inlet and outlet port nozzles 28 are connected to the lower distributor 24 and a high-temperature side water flow passing pipe 30, drooping along the inner wall of the tank main body 20, is connected to the upper distributor 22 while high-temperature side water inlet and outlet port 32 is connected to the lower end of the high-temperature side water flow passing pipe 30.

2. 发明专利

JP2002196030A METHOD FOR DIAGNOSING DETERIORATION OF POWER CABLE 失效
公开(公告)号：JP2002196030A
公开(公告)日：2002-07-10
申请号：JP2000392171
申请日：2000-12-25
申请人： FURUKAWA ELECTRIC CO LTD , TOKYO ELECTRIC POWER CO
发明人： YAGI YUKIHIRO , TANAKA HIDEO , NAKADE MASAHIKO , INOUE DAISUKE
IPC分类号： G01R31/12
摘要： PROBLEM TO BE SOLVED: To diagnose deterioration of a power cable with high accuracy without effect of a cable for a connection even when a measuring site is away from a cable to be diagnosed. SOLUTION: The cable 2 for the connection is connected to the cable 1 to be diagnosed and alternating voltage with a commercial frequency is applied from charging equipment, and a current flowing through insulating materials of the cable 1 to be diagnosed and the cable 2 for the connection is detected by a current transformer 4. Difference between the current detected by the current transformer 4 and a current flowing between a central conductor of the cable 2 for the connection and a shielding conductor, is obtained by a differential amplifier 8a. Difference between the difference and a current flowing through a lossless standard capacitor 5 is obtained by a differential amplifier 8b and a loss current is extracted. Waveforms of the loss current extracted are fetched by a digital oscilloscope 9 as discrete numeric data. A third harmonic component in the loss current is calculated and the deterioration of the power cable is diagnosed by FFT analyzing the discrete numeric data by a computer 10.

3. 发明专利

JPH0843475A METHOD AND APPARATUS FOR DETECTING PARTIAL DISCHARGE IN INSULATING APPARATUS 失效
公开(公告)号：JPH0843475A
公开(公告)日：1996-02-16
申请号：JP18252194
申请日：1994-08-03
申请人： TOKYO ELECTRIC POWER CO , TOSHIBA CORP
发明人： TANAKA HIDEO , OGI HIROMI , OKI ISAO , UCHIDA YOKO , MARUYAMA SHIRO , WAKABAYASHI SEIJI , SAKAKIBARA TAKAAKI , HIRATA KENJI
IPC分类号： G01R31/12 , G01R31/327 , G01R31/333
摘要： PURPOSE:To precisely detect a partial discharge with reference to an unknown partial-discharge waveform without being affected by a noise by a method wherein frequency-characteristic data on the partial-discharge waveform or the like is learned in advance in a neural network. CONSTITUTION:A gas is sealed in a metal container 1 for a gas-insulated apparatus, and a high-voltage conductor is supported by insulating spacers 2a, 2b. A simulation partial-discharge signal is injected into a buried electrode 4b via a signal--injection conductor 12 from a simulation partial-discharge generator 13. Then, regarding frequency-characteristic data on a partial-discharge radio wave obtained by a partial-discharge detector 7 and on a noise waveform measured in a normal operation, features are extracted respectively. The partial- discharge waveform and a noise pattern are learned in advance in a neural network for the detector 7. A discharge waveform which is generated in the apparatus is detected by the learned detector 7, and a plurality of pieces of frequency-characteristic data which are obtained from the detected waveform and whose peak-hold time is different are input to the neural network so as to be judged.

4. 发明专利

JPH07262866A DETECTING METHOD FOR PARTIAL DISCHARGE 失效
公开(公告)号：JPH07262866A
公开(公告)日：1995-10-13
申请号：JP5161894
申请日：1994-03-23
申请人： TOKYO ELECTRIC POWER CO , TOSHIBA CORP
发明人： TANAKA HIDEO , OGI HIROMI , OKI ISAO , UCHIDA YOKO , MARUYAMA SHIRO , HIRATA KENJI
IPC分类号： H01H9/50
摘要： PURPOSE:To perform exact and high-speed partial discharge detection without adversely effected by noise by utilizing the difference in characteristics between the cumulative value distribution pattern of the frequency characteristic data of a partially discharged waveform having the possibility of generation in an insulated apparatus and the cumulative value distribution pattern of the frequency characteristic data of a noise waveform. CONSTITUTION:Partial discharge detection is done in accordance with the following steps (STs): ST1: Generates partial discharge or a noise in a gas insulated apparatus; ST2: to detect its partial discharge waveform or its noise waveform; ST3: to extract the characteristics of its frequency characteristic data from the resulting detected waveform for standardization; ST4: to pattern the cumulative value distribution of frequency characteristic data; to obtain the pattern data of the cumulative value distribution of the detected waveform frequency characteristic data; ST5: to input the resulting data in a neural network; ST6: to determine whether the detected waveform is the waveform generated by the partial discharge or the noise; ST7: to finally display the determined result.

5. 发明专利

JPH05333079A 失效
公开(公告)号：JPH05333079A
公开(公告)日：1993-12-17
申请号：JP14156492
申请日：1992-06-02
申请人： MITSUBISHI ELECTRIC CORP , TOKYO ELECTRIC POWER CO
发明人： IZUI YOSHIO , OGI HIROMI , TANAKA HIDEO , AKIMOTO NORIAKI
IPC分类号： G01R31/00 , G01R31/12 , G05B23/02 , H02H1/00 , H02H7/22
摘要： PURPOSE:To identify the cause of an abnormality as well as the information whether normal or abnormal by providing a neural circuit network device identifying the cause of the abnormality in the case an unknown sensor output is input. CONSTITUTION:If a foreign material 5 comes in and partial discharge 6 is caused, internal insulation gas is vibrated and the vibration propagates to a tank 2. As a result, the tank 2 mechanically vibrates and the vibration is obtained as acceleration change in an accelaration sensor 7 output. A calculator 8 converts it to a spectrum with an FFT and a calculator 9 time-averages it. A calculator 10 normalizes the magnitude and inputs in a neural circuit network device 11. As in the neural circuit network device 11, the relation between the abnormality cause and the spectrum memorized as learned data is stored as each combining weight 19, a neural element corresponding to the abnormality cause outputs a large power based on the memory when an unknown spectrum of the acceleration sensor is input. As the result, not only abnormal or normal is identified but also the generation of the abnormal state corresponding to the label according to the output layer 18b of the neural element which is generated in an electric appliance such as gas insulation breaker, is identified.

6. 发明专利

JPH0522863A METHOD AND APPARATUS FOR MAKING OPERATION PROCEDURE 失效
公开(公告)号：JPH0522863A
公开(公告)日：1993-01-29
申请号：JP16958991
申请日：1991-07-10
申请人： MITSUBISHI ELECTRIC CORP , TOKYO ELECTRIC POWER CO
发明人： TAKANO TOMIHIRO , MATSUMOTO KEINOSUKE , AKIMOTO NORIAKI , TANAKA HIDEO , OGI HIROMI , YOSHIZAWA JUNICHI
IPC分类号： H02J3/00 , G06F9/44 , G06N5/04 , H02J13/00
摘要： PURPOSE:To make an operation procedure with a high speed by a method wherein, if an applicable learning knowledge exists, the operation procedure is made by using the learning knowledge and, if it does not, the operation procedure is made by using searching method and registered as a new learning knowledge. CONSTITUTION:The present state and the target state of a power system are read by a procedure making control function 22a as an operation procedure making subject 21. Then an instruction to judge whether a learning knowledge applicable to the operation procedure making subject 21 exists or not is outputted. A first procedure making means 22b receivers the instruction and judges whether the learning knowledge is registered in a learning knowledge data base 22c or not. If it is recognized that the learning knowledge applicable to the operation procedure making subject 21 is not registered, an instruction that the operation procedure must be made is outputted to a second procedure making means 22d. The second procedure making means 22d receives the instruction and, after it possible the subject with the subject of each bus-bar unit by using the construction knowledge of conception stored in a domain knowledge data base, makes an operation procedure 24 by using a searching method.

7. 发明专利

JPH01290062A NERVE SIMULATION CIRCUIT NETWORK DEVICE 失效
公开(公告)号：JPH01290062A
公开(公告)日：1989-11-21
申请号：JP11963288
申请日：1988-05-17
申请人： TOKYO ELECTRIC POWER CO , MITSUBISHI ELECTRIC CORP
发明人： TANAKA HIDEO , MATSUDA SEI , OGI HIROMI , IZUI YOSHIO , TAOKA HISAO , SAKAGUCHI TOSHIAKI
IPC分类号： G06F15/18 , G06G7/60 , G06N3/00 , G06N3/04 , G06N99/00
摘要： PURPOSE:To obtain an answer of an optimization problem regardless of some troubles of the nerve simulation elements and the connection elements by allocating the approximately same working states to the problems to be solved by the nerve simulation element groups for multiplexing and also multiplexing the connection element groups which feed back the output signals of the nerve simulation elements. CONSTITUTION:Each nerve simulation element contained in the nerve simulation element groups 11 has a nerve simulation circuit network 1 which calculates the sum total of the feedback signals received via each connection element 21 and the independent signals 31 approximately equal to each other and inputted separately from those feedback signals and at the same time secures locally a minimum energy state for an aggregate. Furthermore, a computer 3 is added to perform the transfer of information to the network 1. Thus it is possible to obtain an answer of a combined optimization problem in a stable way regardless of some troubles of the nerve simulation elements and the connection elements.

8. 发明专利

JPS60207425A DEVICE FOR INTERLOCKING POWER DISTRIBUTION SYSTEM 失效
公开(公告)号：JPS60207425A
公开(公告)日：1985-10-19
申请号：JP6196584
申请日：1984-03-29
申请人： TOKYO ELECTRIC POWER CO , MEIDENSHA ELECTRIC MFG CO LTD
发明人： MORI SADAO , YOROZUYA TSURUO , KOBAYASHI TADAO , WATANABE NORIYUKI , TANAKA HIDEO
IPC分类号： H02J3/06

9. 发明专利

JPH0476685A NEURAL CIRCUIT NETWORK DEVICE 失效
公开(公告)号：JPH0476685A
公开(公告)日：1992-03-11
申请号：JP18649790
申请日：1990-07-12
申请人： MITSUBISHI ELECTRIC CORP , TOKYO ELECTRIC POWER CO
发明人： AKIMOTO YOSHIAKI , TANAKA HIDEO , OGI HIROMI , IZUI YOSHIO , TAOKA HISAO , SAKAGUCHI TOSHIAKI
IPC分类号： G06G7/60 , G06F15/18 , G06N3/08 , G06N99/00
摘要： PURPOSE:To shorten an execution time due to learning by setting up values obtained by respectively multiplying the input value of learning data and a value calculated from frequency by an output value and a prescribed coefficient as coupling weight between an intermediate layer and an output layer. CONSTITUTION:This neural circuit network device is constituted of an input layer 11a, an intermediate layer 11b and an output layer 11c and provided with plural nerve elements 11 simulating nerve cells in an organism and coupling weight for coupling between nerve element layers. Then frequency for the learning data constituted of a pair of an input value obtained at a fixed interval and an output value corresponding to the input value is set up as coupling weight for coupling the input layer and the intermediate layer by using a procedure for developing a function forming the learning data by a sampling function, developing the sampling function by a triangular function and approximating the triangular function by a sigmoid function, a value calculated by the frequency and the input value of the learning data is set up as the bias of the intermediate layer, and a value obtained by the frequency and the input value and output value of the learning data is set up as coupling weight for coupling the intermediate layer and the output layer to rapidly obtain a required output.

10. 发明专利

JPH0251326A STOP SCHEDULE ADJUSTMENT ASSISTING DEVICE 失效
公开(公告)号：JPH0251326A
公开(公告)日：1990-02-21
申请号：JP20059788
申请日：1988-08-11
申请人： TOKYO ELECTRIC POWER CO , TOSHIBA CORP
发明人： AKIMOTO YOSHIAKI , TANAKA HIDEO , OGI HIROMI , TAKESHIMA YURIKO , SHINOHARA JUNICHI , KUNUGI MASAHIKO , NATSUMURA ETSUROU , ASAI TOSHIO
IPC分类号： H02J3/00
摘要： PURPOSE:To carry out adjustment accurately and quickly by reasoning the relation between the influence of stop schedule onto a power system and another stop schedule then displaying the validity of the stop schedule. CONSTITUTION:A stop schedule data storing means 11 stores the content of a stop schedule while a power system date storing means 12 models a power system and stores the modeled state and a knowledge storing means 13 stores knowledge necessary for judgement of the validity of the stop schedule. Upon request from a CRT unit 2, a reasoning means 14 applies the knowledge stored in the knowledge storing means 13 for the content stored in the stop schedule data storing means 11, and reasons the influence onto the content stored in the power system data storing means 12 or the relation between the stop schedule data storing means 11 and other stop schedule. Then the validity of the stop schedule is displayed on a CRT unit 2.

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