Research Of Core Power Reconstruction Using Disturbances Search Method Based On Out-Core Detection

The on-line monitoring of nuclear reactor power distribution can improve economic performance while ensuring the safety of nuclear power plants.For the second generation pressurized water reactors,the in-core measurement system can only periodically measure the core power through some components and cannot achieve the function of on-line power distribution monitoring.While,the out-core nuclear instrument system can achieve thus function by measuring the neutrons leaked from the reactor.The difficulty in reconstructing the core power distribution based on out-core measurements is to reconstruct a complex core power distribution by using limited detector data.By investigating the existing PWR power distribution reconstruction method,the current methods mainly include transmission matrix method,harmonic synthesis method,and neural network method.The transmission matrix method is a linear method and the accuracy is not enough.Harmonic synthesis method does not take into account changes in cross-sections under varying operating conditions when solving each order of harmonics,and does not consider the specific reasons for the influence of various physical factors on power distribution changes.The neural network method completely ignores the diffusion equation and k_eff and other physical constraints,and cannot reasonably explain the possible deviations.This project is to find a method to analyze the influence of various physical factors on the section,based on the influence of the section deviation on the power distribution change,and calculate the physical constraints,in addition to the final power distribution.The reason for the power shape change can also be obtained,so that the core power reconstruction result is more reliable.The method establishes the relationship between the disturbance factor causing the power distribution change and the data change of the external detector.The CSA continuously corrects the disturbance variable according to the deviation between the actual data of the detector and the calculated data until the above deviation is zero.At this point,the core power distribution under the influence of the disturbance factor is the true core power distribution.In this paper,the factors affecting the core power distribution are firstly modeled.Secondly,the feature statistical algorithm program and the second type of boundary condition Green's function block method diffusion program need to be rewritten for the problem.Finally,the temperature fluctuations and xenon poison are studied as the influencing factors,and the temperature fluctuation calculation examples are constructed.The examples of xenon poison oscillating and the simultaneous occurrence of temperature fluctuations and xenon poison oscillating are used to verify the feasibility and accuracy of the proposed method.The temperature fluctuation-xenon poison oscillating example shows that the relative power deviation of the core can be reduced from 6.57%to 0.245%.The data show that the method can successfully search for the influencing factors and core power distribution,and the feasibility of the method is preliminarily verified.