Font Size: a A A

Research On Distributed Condition Monitoring Method For Primary Loop Of Nuclear Power Plants

Posted on:2016-12-13Degree:MasterType:Thesis
Country:ChinaCandidate:F XuFull Text:PDF
GTID:2322330542973950Subject:Nuclear science and engineering
Abstract/Summary:PDF Full Text Request
Security is the most important factor in nuclear power plant.Nuclear power plant safety issues can be considered from two aspects,one of which is the safe operation of the systems and equipment,and the other is the safe operation of the sensor.It will cause unnecessary downtime or even more serious problems,if we couldn't detect and distinguish anomalies timely which occur in the nuclear power plant equipment or sensors.Therefore,there is a great practical significance in monitoring nuclear power plant equipments and sensors.The paper aims to study distributed condition monitoring methods,which is detailed as follows:(1)A research was made for the primary system and auxiliary system for primary loop of Qinshan nuclear power plant,including reactor coolant system(RCS),chemical and volume control system(CVCS),component cooling water system(CCWS),residual heat removal system(RHRS).Based on distributed strategy,the whole monitoring task of sensor and equipment is divided into different subunits.(2)With principal component analysis,auto-associative kernel regression,sequential ratio probabilistic,monitoring models were built for grouped sensors.With these multiple models and the correlative characteristic of the measured datum,we can monitor,identify and reconstruct faulty sensors.(3)The research objects are divided into two-level subunits,system/equipment monitoring subunits.The system subunits use multiple PCA models to monitor operating condition of NPP and detect faults,and then it can provide entry condition for equipment monitoring subunits.The equipment monitoring subunits use physical equations to modeling,and use Expert System to replenish those models when it is hard to modeling or there is a much easier way to analysis.Critical safety functions are used as principles of conclusion fusion to get the optimal solution.(4)On the basis of research mentioned above,a NPP Distributed Condition Monitoring System is developed,and tested on full scope nuclear power plants simulator.Research results show that the distributed condition monitoring system can identify and reconstruct faulty sensors effectively,and monitor the nuclear power plant systems and equipment validly.In summary,the whole condition monitoring system is reliable andaccurate,and it also lays a technical foundation for the future research.
Keywords/Search Tags:Sensor Calibration Monitoring, Equipment Monitoring, Principal Component Analysis, Auto-associative Kernel Regression, Physical Modeling
PDF Full Text Request
Related items