Reasearch On Error Analysis And Compensation Techniques Of Reactor Power Monitoring System

Nuclear power has already become a significant pillar in the energy development and structural adjustment in China. The reactor power monitoring system is an organic part of the reactor instrument & control system,and to a certain extent,the monitoring results may impact the safety and economic operation of the entire reactor system. The data measurement of the reactor power monitoring system is realized with the assistance of various measuring device. Errors in the measuring process may impact the operation of nuclear power plant.Therefore, it is quite necessary to analyze and compensate for the errors in the reactor power monitoring system.In this paper, it attempts to seek for an error compensation techniques based on intelligent sensors in the digital instrument & control system through the studies on the reactor power monitoring system and error compensation techniques.The monitoring principles of two reactor power monitoring systems is studied, and various factors that may impact the generation of errors is found out by applying the error propagation theory for analyzing the errors' propagation in the measurement system.Furthermore, it is concluded for the basic principles of compensation during the measuring process by aiming at different measuring approaches, and the targets of compensation are determined. On that basis, neural network theory is applied into the compensation of errors in the reactor power monitoring system. Moreover, a neural network with certain self-organization ability is applied to the error compensation, and the results of error compensation with such techniques are evaluated.Based on the analysis of digital instrument & control system in nuclear power plants and its technical features, a solution based on intelligent sensors is designed for achieving neural network compensation. Furthermore, corresponding intelligent sensor is designed for verifying the validity of the compensation solution.The achievements of this paper provides certain theoretical supports for improving the precision and program of reactor power monitoring system and enhancing the economy of nuclear power plant.