In the field of nondestructive inspection by eddy current testing, various approaches to the depth-sizing of SCCs by numerical inversion analysis have been taken so far. However, its accuracy is not so high and limited to a certain extent because of no information about the real value of inner conductivity of SCCs. Therefore, we first investigated experimentally the inner conductivity distribution of several artificial SCCs by a precise mechanical processing and 4 terminal resistance measurement. From ...
英字タイトル:
Improvement of a SCC depth-sizing technique using ECT analysis with detailed investigation of SCC conductivity distribution

Our Company, aims at tracking huge plant parameters (Big Data), managed by Nuclear Power Plants. By application of this Big Data, along with NEC Corporation we developed a system to detect any equipment abnormalities accurately and as much as possible at an early stage. After the implementation of the system, checks of the actual working of the system can be conducted and performance gains enhanced drastically. ...
英字タイトル:
Development of the Plant Monitor System utilized Invariant Analysis Technology and Big Data

The experiments that imitate wear in pump impellers were performed, in which various physical values such as outlet pressure, suction pressure, flow rate and motor current were acquired simul taneously. Three impellers with various wear amount and a normal one were tested. Similarity Based Modeling (SBM) was applied to the data for detection of the abnormality. SBM is a non-parame tric modeling method that makes a model of a certain state by sampling vectors that consists of various physical values at the ...
英字タイトル:
Investigation of multivariate analysis for detection of wear in pump impellers

A method that applies the neural network to acoustic data to detect abnormalities of horizontal pumps is proposed. In the experiments, running sound of a pump with bearing damage or misalignment was acquired from various distance while another pump in the same room was running or not running as a possible noise source. By applying the neural network to the appropriately processed frequency spectrum of acquired data, abnormality can be accurately detected regardless of the distance between the sensor and ...
英字タイトル:
Detection of abnormality of pumps by acoustic diagnosis using neural network

Rationalization of the plant maintenance is one of the main topics being investigated in Japanese nuclear power industries. Optimization of the inspection and replacement period of equipments is effective for the maintenance cost reduction. The more realistic model of the replacement policy is proposed in this study. It is based on the classical replacement policy model and its cost is estimated. Then, to consider the inspection for the maintenance, the formulation that includes the risk concept is d...
英字タイトル:
Optimization on Replacement and Inspection Period of Plant Equipment

A new electromagnetic method for the nondestructive evaluation of electric pumps with ball bearings is proposed. A permanent magnet generates a static magnetic field and a coil of several hundred turns wrapped around it is used to pick up eddy currents induced inside rotating ball bearings and impellers. First, a method that makes use of the autocorrelation is applied to the signals of the ball bearings. A relation between the processed signals and the size of the defects in the bearings is obtained....
英字タイトル:
Electromagnetic Nondestructive Detection for Wounded Bearings and Correlation of Size of Defect and Signal

We have developed the prediction monitoring and diagnostic system for the purpose of the early detection of the abnormal behavior of the complicated system like a nuclear power plant. Furthermore, the techniques for the modeling of the relationships between sensors are applied to evaluate the system performance. The developed method is applied to the primary loop recirculation (PLR) pumps and the system performance indicator is calculated. The results are compared with the maintenance record of the PLR p...
英字タイトル:
Application of Prediction Monitoring and Diagnostic System to Plant Performance Evaluation

It is thought that we can let knowledge of "Human" substitute by building a function to utilize predictive information, and to extract a similar example from these documents in AI. In this way, we can largely reduce a burden on "Human" by building past accidents as a database, and being resemblance or equivalence, and extracting a cormorant from that. But we may list it by expression varying among document creators even if it is the same company. When we performed a document search with disregard to such a ...
英字タイトル:
Examination of a method to search past similar cases registered in NUCIA from predictive information detected by SIAT

System Invariant Analysis Technology, SIAT can detect abnormal behavior of plant parameters. However, the cause analysis functionality for the abnormal behavior has been strongly desired to support engineers making decisions for taking appropriate actions. For that purpose, we have developed the cause analysis functionality based on the result of the broken invariant analysis. Database of the broken invariant distribution with that cause is prepared and the similarity with the current abnormal behavior of t...
英字タイトル:
Development of the cause analysis functionality for the abnormal behavior detected by SIAT

A place to depend on an experienced of the knowledge of operation engineer and the maintenance engineer so-called experienced engineer about plant administration is big and, in us, will entrust it to "Human" from a process of the phenomenon identification to the last judgment of the result. Because a human resource is limited, we cannot depend on "Human" for all the judgments. Therefore, about keeping a function to substitute in [Human = experienced person] in AI namely SIAT and other AI system namely IBM W...
英字タイトル:
Advance of plant monitoring technology by cooperation of an system invariant analysis technology (SIAT) and IBM Watson Explorer