Optimization Of Scatterers Geometric Model And Reduction Of CCD Image Noise In The Measurement Of Wet Steam Parameters

At present,with the increasing number of large-capacity high-parameter turbine units,higher requirements are placed on the economical efficiency and safety of power plant operation.Humidity of turbine internal steam is an important factor affecting the efficiency and safety of the turbine.To guide the operation of the turbine,it is of great significance to perform real-time,accurate and non-contact measurement of parameters of steam of turbine internal.This paper draws lessons from domestic and foreign research experience,designed the back-angle scattering method to measure the steam humidity based on the Mie scattering theory,and optimized its geometric model.Using a CCD camera as a data acquisition device,an experiment is established based on an off-axis angle scattering method.The platform carried out experimental research and optimized the experimental signals.Firstly,on the basis of analyzing the measurement model of the off-axis scattering method,the geometric model of the scatterer is constructed and optimized according to the knowledge of light scattering and the CCD imaging principle.The volume of scatterer and the parameters related to the scatterer are calculated.Then,the mathematical expression of volume of the optimized geometric model is substituted into the measurement model,and the measurement model is simulated using Matlab software.The theoretical value of the scattered light intensity is obtained and compared with the geometric model of the scatterer before optimization and analysis.The results show that the relative error between theoretical and experimental values of scattered light intensity is reduced after optimization of the geometric model of scatterers,and the change trend is more consistent,which verifies the superiority and accuracy of geometric model optimization.Then,on the basis of comparative analysis of various classical filtering algorithms,three filtering algorithms,namely mean filtering,wavelet threshold and Wiener filtering,are selected.According to the characteristics of noise of CCD images and the selected algorithm characteristics,the wavelet threshold denoising and Wiener filtering improves.Using Matlab software,the SNR and RMSE are used as the evaluation criteria.The filtering results of the same image before and after the improvement of several filtering algorithms are compared.The results show that the improved Wiener filter with 3×3 window has the best effect of noise reduction.The SNR of the noise-reduced image is 22.35595 and the RMSE is 0.4082,which are the optimal values in several filtering algorithms.This article chooses the improved filter algorithm to denoise the experimental images.Finally,the uniform search is used to invert the data obtained from the experimental measurement,and the wet steam parameters are obtained and the humidity is calculated.The humidity variation law under the same working condition and the variation law of the humidity under the different working conditions are analyzed.The analysis results show that the calculated results of humidity under the same conditions are 1.686%,1.684%,and 1.718%,respectively.The results are very close to each other and are in line with the actual conditions.The relationship between the cross-section 3 and cross-section 4 back pressure and humidity under variable conditions is also consistent with the principle of thermodynamics.From the side to prove that this measurement system is more reliable for measuring wet steam humidity.