用于两相流测量的ECT图像重构技术研究

电容层析成像技术(ECT)具有非侵入、响应速度快、成本低等优点,是用于两相流参数检测非常有发展潜力的技术之一。而图像重构是ECT系统研究的关键技术。该文利用有限元方法对12电极ECT系统进行建模仿真,进行正问题求解,获得了图像重构的样本数据;引入改进的径向基函数神经网络,建立了ECT图像重构算法,并在MATLAB平台上进行了仿真验证。结果表明,改进的径向基神经网络算法在图像重构准确度及速度方面有了明显提高。

An ECT image reconstruction algorithm in two-phase flow measurement

Due to its many distinct advantages such as no invasion,high speed and low cost,electrical capacitance tomography is becoming a very promising technology in two-phase flow process parameters measurement.Image reconstruction algorithm is an important factor of ECT system.An 12-electrodes electrical capacitance tomography system is studied in forward and inverse aspects,respectively.The capacitance values are obtained by the finite element（FE） model of the sensors using ANSYS,and they are used for the sample data of the image reconstruction.An improved radial basis function（RBF） neural network algorithm is used in ECT image reconstruction.The algorithm is emulated and validated on MATLAB.Simulation results show that the algorithm has reconstruction quality better than that of LBP and BPNN algorithm,and reconstruction time shorter than that of the LBP,BPNN and original RBFNN algorithm.