Research On On-line Monitoring Method And System Of Small Modular Reactor

Three dimensional reactor core power distribution monitoring is an effective way to fully improve the economy of nuclear energy while ensuring the safety of nuclear reactor. Small modular reactor is a new distributed energy system, and has many features, such as short construction period, small investment and flexible installation. This thesis is focused on the research of on-line monitoring method and system of a small modular reactor ACP100, the main research contents include:Firstly, the pre-processing methods of detector signal are studied. This part is consist of the research of the dynamic compensation method of self powered neutron detectors(SPND) based on robust filter and the research of the fault detection method of SPND based on principal component analysis(PCA). The response time of SPND s can be reduced significantly when the robust filter is used in dynamic compensation, and the balance between response time and noise gain can be obtained by tuning relevant parameters. The PCA model can be constructed based on the mutual relationships of SPND signals, and the features of detector failure, such as the detection and isolation of faulty detector, can be extracted from the PCA model.Secondly, the control rod position monitoring method is studied. The unacceptable od misalignment may occur as a result of a failure of traditional rod positioning system based on eletromechanical devices, and the core monitoring system should has the ability to detect the rod misalignment. The radial basis function neural network(RBFNN) is utilized to build the direct relationship between the signal of SPNDs and the rod position, and the robustness of rod position monitoring can be improved by using the regularized RBFNN. In addition, two correction strategies of traing data sets are proposed to improve the accuracy of position monitoring when there are mismatches between actual physical factors and modeled physical factors.Thirdly, the core power distribution reconstruction methods are studied. The main contents include:(1) The Ordinary Kriging model is researched to reconstruct the whole core power distribution and determine the optimal detector location strategy based on limited incore detectors;(2) The mathematical model of power reconstruction based on excore detectors is analyzed, and the regularized RBFNN method is utilized to reconstruct the core axial power distribution based on excore detector signals;(3) For ACP100, the data assimilation method is used to qualitively analyze the reconstruction result of the combination of incore measurement and excore measurement. A least squares coupling coefficient(LSCC) method which can use the incore measurement and excore measurement integratedly is proposed, and the effectiveness of LSCC method is verified based on the numerical simulation of ACP100.Lastly, the power distribution monitoring system SMROMS for ACP100 is developed, and the monitoring uncertainty of SMROMS is analyzed based on Latin Hypercube Sampling(LHS) method.