Construction Of MCSA Soft Sensing Mathematical Model For Nuclear Power Pump And Research On Hydraulic Performance Monitoring

Digital nuclear power is a major trend in the development of modern nuclear power.It integrates digital technology with nuclear power technology,achieving digitalization,intelligence,and automation of nuclear power systems,and improving the safety,economy,and reliability of nuclear power systems.Nuclear power pumps are an important facility in nuclear power plants,widely used in various areas of nuclear power plants.The efficient operation of nuclear power pumps is a crucial condition for the safe and stable operation of nuclear power plants.As a part of digital nuclear power,the operating conditions of nuclear power pumps require intelligent monitoring.The traditional methods for monitoring the working conditions of nuclear power pumps are mostly invasive,requiring sensors to be installed around or even inside the pump.However,due to safety concerns,these sensors cannot be installed,resulting in traditional monitoring methods being unable to provide technical support for digital nuclear power.Soft measurement establishes a mathematical model between input variables and output variables,finds the mapping relationship between them,and achieves online real-time estimation of target variables.Compared with traditional methods,soft measurement has advantages such as high security,low cost,and simple maintenance.This article proposes a soft sensing technology that relies on the internal flow characteristics of nuclear power pumps and MCSA technology to indirectly monitor the operating conditions of nuclear power pumps.This article adopts theoretical analysis,numerical simulation,and experimental verification methods to study the model of nuclear power pumps.Multiple signal processing methods are applied to the motor stator current signals under different operating conditions,exploring the applicability of relying on the motor stator current to measure the operating conditions of nuclear power pumps,and providing technical support for digital nuclear power.The research content and innovative points of this article are as follows:(1)A systematic comparative analysis was conducted on the monitoring methods for the operating conditions of nuclear power pumps,and a preliminary summary was made on the methods suitable for monitoring the operating conditions of nuclear power pumps.Elaborate on the different research methods used for the internal flow characteristics of different parts of nuclear power pumps,and summarize the application of soft sensor modeling in monitoring the operating conditions of nuclear power pumps.The results indicate that studying the internal flow characteristics of nuclear power pumps helps to understand the internal flow mechanism of nuclear power pumps;Summarized the soft sensing modeling methods and found a suitable soft sensing modeling method for monitoring nuclear power pumps,providing reference for subsequent research.(2)In order to explore the mechanism and feasibility of soft sensing technology for the internal flow characteristics of nuclear power pumps,CFD numerical simulation method was used to compare the internal flow characteristics of different spatial points,blade surfaces,and the entire water body in the impeller flow channel of nuclear power pump models with rotational speeds of 56,77,120,151,134,and 202 under different operating conditions.The research results indicate that the main frequency of pressure pulsation at spatial points in the impeller channel and on the blade surface is the axial frequency;The main frequency of the pulsation signal induced by the impeller flow channel is the blade frequency.(3)In order to construct a monitoring model for the hydraulic performance of nuclear power pumps,a steady-state simulation model of the nuclear power pump device system was established using the simulation theory of the nuclear power pump device system.Based on this,simulation calculations were conducted;A mathematical model for the basic geometric dimensions,flow rate,and head relationship of nuclear power pumps was constructed based on the theorem of moment of momentum,water torque equation,velocity triangle,and energy equation.Combining internal flow characteristics with steady-state simulation mathematical models,provide effective constraint intervals for the construction of MSCA soft sensing mathematical models..(4)In order to obtain accurate and reliable testing devices and methods,a testing platform for soft sensing performance testing of nuclear power pumps was designed and built,and the external characteristics test of nuclear power pumps and the collection of three-phase motor stator current signals were completed.Analyze the systematic,random,and comprehensive uncertainties of the experiment,and based on this,conduct transient uncertainty analysis on current signals under different operating conditions and sampling times.The research results indicate that the error in this experiment is relatively small,and the maximum error under rated operating conditions is around 0.5%.(5)To achieve soft measurement of the hydraulic performance of nuclear power pumps based on internal flow characteristics and MCSA technology,time-domain and frequency-domain analysis of the collected raw current signal is performed,and then combined with SVD to analyze the frequency domain of the raw current signal;Further use wavelet analysis and ITD decomposition methods for time-domain and frequency-domain analysis to obtain the trend of current signal changes under different flow rates;Then,combined with modeling methods based on internal flow characteristics,a hydraulic performance monitoring model for nuclear power pumps is constructed.The research results indicate that the soft measurement accuracy of hydraulic performance of nuclear power pumps based on internal flow characteristics and MASC technology can reach 95%.