【摘要】 针对现有图像重建算法以定性为主,所引起的相浓度参数测量精度有限的问题,该文提出了在对电容层析层像系统传感器测量数据进行分析时,采用主成份分析(PCA)结合人工神经网络反向传播(BP)算法,通过训练神经网络,得到从电容测量值到相浓度之间的对应关系。样本测试可以得到很好的结果,网络输出浓度与设定浓度之差可以控制在±5%范围之内。证明该方法测量两相流相浓度参数是可行的。更多还原

【Abstract】 Electrical capacitance tomography (ECT) is a technique for twophase flow measuring. The phase concentration can be extracted from reconstructed tomograms. For the nonlinearity and illposedness of ECT image reconstruction, the accuracy of phase concentration estimated from image reconstruction is limited. This paper introduced a method based on principal component analysis combined with artificial neural networks for twophase flow concentration measurements. The phase concentration was determined from the raw capacitances data. Simulation results show that the error in the estimated concentration is within ±5%.更多还原